Every taxpayer using ZIMRA's Self-Service Portal (SSP) carries a history that is older than the portal itself. Before TaRMS went live, returns were filed on the legacy e-services platform (or on paper), taxpayers were identified by the legacy Business Partner Number (BPN) rather than the Taxpayer Identification Number (TIN), and assessments and payments lived in ZIMRA's older systems. The Old Period Documents page — the fourth of the five pages in the Tax Return Management module — is where that history surfaces inside the SSP: in the local guide's words, it holds "historical returns from periods predating the SSP, where ZIMRA has migrated them in." That single sentence is the entire procedural description the available source gives of the page, and this lesson says so honestly: the screen-level mechanics carry verification flags throughout, while the legal architecture — which is fully confirmable from the source Acts — carries the weight.

That legal architecture matters because a system migration does not reset a single legal obligation or right. The lesson's core doctrine is the six-year lattice: four distinct six-year clocks, each anchored to a different starting date, govern what can still happen to an old period. Section 37B of the Income Tax Act [Chapter 23:06] (mirrored by Section 57(3) of the VAT Act [Chapter 23:12]) obliges the taxpayer to retain books and records for 6 years from the date of the last entry. Section 47 of the Income Tax Act lets the Commissioner reopen an old year within 6 years of the relevant assessment — and without any time limit at all where there has been fraud, misrepresentation or wilful non-disclosure. Section 41 of the VAT Act bars the Commissioner from recovering unreturned VAT after 6 years from the date the amount became payable — but only where the taxpayer proves three cumulative good-faith conditions, and only if no assessment was issued inside the window; the Triangle/Hippo Valley litigation shows the bar is lost if not raised. Finally, Section 48 of the Income Tax Act and Section 44 of the VAT Act cap the taxpayer's own refund claims at 6 years (with a brutal 6-month fuse where the original payment followed the practice generally prevailing).

The practical consequence is that the Old Period Documents page is not nostalgia — it is evidence. Migrated historical returns are ZIMRA's record of what was declared in the pre-SSP era; your own retained records under Section 37B are your record; and an audit, an objection, a refund claim, a due-diligence exercise or a deceased-estate wind-up may turn on reconciling the two. Section 42 of the VAT Act makes a document produced by the Commissioner purporting to be a copy of a notice of assessment conclusive evidence of the assessment (except on appeal), which raises the stakes of checking the migrated record early, while the underlying papers still exist.

This lesson distinguishes old-period work from its neighbours: back-filing (unfiled SSP-era periods cured through Pending Tax Returns — the previous lesson) and amendment (wrongly filed SSP-era returns — two lessons back). It maps the three lanes, walks the four six-year clocks clause by clause, and builds a reconciliation routine for the first time a taxpayer opens the migrated archive. Procedural specifics of the page itself — whether returns for pre-SSP periods can be initiated from it, its search filters, its export options — are not confirmable from the available sources and are flagged accordingly.

Every revenue administration that modernises its systems faces the same problem: the new system starts on a particular day, but taxpayers' legal histories do not. On the day TaRMS went live, every registered taxpayer in Zimbabwe carried — into a brand-new system — years of filed returns, assessments, payments, credits, disputes and unresolved balances generated on the legacy e-services platform and its predecessors.

The law took no notice of the migration. The Income Tax Act [Chapter 23:06] and the VAT Act [Chapter 23:12] speak of years of assessment, tax periods, returns, assessments and records — never of platforms. An assessment raised in the legacy era remains an assessment; a record-retention period that began under the old system keeps running under the new one; a refund window that opened before go-live keeps closing on its own schedule. The system is new; the periods are not.

The SSP acknowledges this in the Tax Return Management module. As established in the Return Submission lesson, the module has five pages: Pending Tax Returns, Submitted Tax Returns, E-Agreement, Old Period Documents, and Drafts. The local guide describes the Old Period Documents page in one sentence:

"Old Period Documents — historical returns from periods predating the SSP, where ZIMRA has migrated them in."

That is the entire confirmed procedural description. It establishes three things with reasonable certainty: (1) the page concerns periods predating the SSP — the legacy era; (2) its contents are returns, i.e. declarations actually filed in that era; and (3) those contents were migrated in by ZIMRA, not created by the user. Everything beyond that — search filters, whether documents can be exported, whether a new return for a pre-SSP period can be initiated from this page or must be routed elsewhere — is unconfirmed and flagged where it arises.

Why does a lesson-length treatment of a one-sentence page matter? Because old periods are where some of the most expensive events in a taxpayer's life happen:

• Audits reach backwards. Section 47 reassessment reaches six years back as of right — and indefinitely on fraud, misrepresentation or wilful non-disclosure. An audit opened today routinely interrogates periods that were filed on the legacy platform.
• Refunds expire silently. The six-year (and sometimes six-month) refund windows in Section 48 and VAT Section 44 keep running whether or not anyone is watching. Money recoverable from an old period is lost by lapse of time, not by refusal.
• Evidence decays. Businesses lose servers, change accountants, destroy archives. The migrated record in Old Period Documents may eventually be the only accessible record of what was declared — for better or worse.
• Transactions interrogate history. Due diligence on a company purchase, a bank's Section 60B credit check, an executor winding up a deceased estate, a liquidator reconstructing a company's affairs — all reach into pre-SSP periods.

This lesson completes the returns arc's treatment of time. The Return Submission lesson taught the current cycle; Amending a Current Return taught correction of what was filed; Back-Filing Prior Returns taught curing what was never filed in the SSP era. Old-Period Returns teaches the third lane: periods that predate the system itself.

No provision of either Act deals with "old periods" as such — the category is created by the system cutover, not by statute. What the statutes supply is a set of period-based rules that apply identically to pre-SSP and SSP-era periods. Five clusters matter.

B.1 The record-retention obligations: ITA Section 37B and VAT Section 57(3)

Section 37B(1) of the Income Tax Act (substituted by Act 3/2009 w.e.f. 23 April 2009) provides, confirmed verbatim from the source Act:

"Every person whose gross income does not consist solely of salary, wages or similar compensation for personal service, shall keep or cause to be kept in the English language, proper books and accounts of all his or her transactions and, unless otherwise authorised by a competent court or by the Commissioner, shall retain for a period of 6 years from the date of the last entry therein all ledgers, cash-books, journals, paid cheques, bank statements and deposit slips, stock sheets, invoices, and all other books of account relating to any trade carried on by him or her and recording the details from which his or her returns for the purposes of this Act were prepared."

Unpack the clause:

• Who: everyone except pure employees (persons whose gross income is solely salary/wages). Every trader, company, landlord, professional and mixed-income individual is caught.
• What: the enumerated list is wide — ledgers, cash-books, journals, paid cheques, bank statements, deposit slips, stock sheets, invoices — and then closes with a catch-all: all other books of account recording the details from which returns were prepared. The test is functional: if a return was prepared from it, retain it.
• How long: 6 years from the date of the last entry in the book — not from the filing date, not from the year-end. A ledger whose last entry was made on 15 March 2024 must be kept until 15 March 2030, even though it records transactions stretching back earlier.
• Language: English. A record kept only in another language does not satisfy the section.
• Escape: only a competent court or the Commissioner can authorise earlier disposal.

Section 37B(2) makes contravention an offence punishable by the greater of a level 7 fine or a fine equal to 10% of the person's taxable income, or imprisonment up to three months, or both (penalty clause substituted by Act 8/2011 w.e.f. 1 January 2012). The 10%-of-taxable-income alternative is the sting: for a profitable company, a records failure can dwarf a level 7 fine. The source Act annotates the section with NYS v ZIMRA 19-HH-617.

The VAT mirror is Section 57(3), confirmed verbatim: records must, "where kept in book form, be retained and carefully preserved by the registered operator for a period of 6 years from the date of the last entry in any book," or, where not in book form, "for a period of 6 years after the completion of the transactions, acts or operations to which they relate." Section 57(4) lets the Commissioner authorise retention of the information* in an alternative acceptable form in lieu of originals — except for ledgers, cash books, journals and paid cheques, which must be kept as they are. The source Act annotates the failure point with PIL (Pvt) Ltd v ZIMRA 17-HH-213*.

Finally, recall from the Roles & Permissions lesson that Part VIIIA modernises the medium: Section 80H allows electronic retention to satisfy Section 37B, and Section 80C makes electronic data admissible with due weight (PIL again). A scanned, well-indexed archive of the legacy era is legally as good as the paper — and practically far better.

B.2 The Commissioner's reach backwards: ITA Section 47 and VAT Section 31

As walked verbatim in the Amending a Current Return lesson, Section 47 of the Income Tax Act obliges the Commissioner ("shall adjust") to raise an additional assessment where income was omitted, a loss overstated or a credit wrongly allowed — subject to two shields: proviso (i), where the original treatment followed the practice generally prevailing at the time, and proviso (ii), a six-year bar running from the relevant assessment — "lifted entirely" where the under-assessment is due to fraud, misrepresentation or wilful non-disclosure. The annotated case line ('T M Fee', A v COT 00-ITC-1691, Deb 19-HH-664, Man 20-HH-078, Bath 20-HH-552, M Safaris 20-HH-331, IAB 22-HH-032, Zimplats 21-SC-159) was set out there and is not re-walked here; the point for this lesson is its temporal consequence: a legacy-era year is reachable today if it falls within six years — or forever, if the taxpayer's conduct in that era crossed the line. Section 47(2) imports the Section 45–46 machinery, so a reopened old year carries the full additional-tax exposure.

On the VAT side, Section 31(3)(b) lets the Commissioner assess where not satisfied with any return — including a legacy-era return — and Section 31(4) permits estimation. But VAT adds something income tax does not have, and it is the heart of the next cluster.

B.3 The VAT recoverability bar: Section 41, confirmed verbatim

Section 41 of the VAT Act contains a taxpayer-protective limitation unique in the two Acts. Where, in consequence of any of the following —

"(i) any amount of tax chargeable under this Act in respect of a supply of goods or services has not been returned in any return required to be furnished under section twenty-eight or twenty-nine …; or (ii) excise duties are imposed in terms of the Customs Act; or (iii) any amount of tax chargeable under this Act in respect of a supply of imported services has not been accounted for and paid as required by subsection (1) of section thirteen; or (iv) any amount of tax has been incorrectly deducted in terms of paragraph (3) of section fifteen in any return …"

— an amount of tax which should have been paid was not paid, then:

"that amount shall not be recoverable by the Commissioner after the expiration of a period of 6 years reckoned from the date on which that amount became payable in terms of this Act, if it is shown— A. that the failure to pay the amount which should have been paid was not due to an intent of the person concerned or any other person under the control or acting on behalf of that person not to make payment of tax; and B. that the person responsible for the payment … acted in good faith and on an assumption that an exemption or a rate of zero % was in fact applicable …, or that any such supply was not subject to tax under this Act, … or that a deduction … was in fact applicable …; and C. that the said assumption was based on reasonable grounds and not due to negligence on the part of the said person: Provided that this paragraph shall not apply if the Commissioner has not later than the end of the said period issued an assessment in respect of the unpaid tax."

Read the structure carefully, because every word allocates risk:

1. It is a recoverability bar, not an innocence presumption. The tax was chargeable; the section only stops the Commissioner recovering it after six years from the date it became payable.
2. The conditions are cumulative and the onus is the taxpayer's ("if it is shown"). No intent to avoid payment (A), and good faith on a specific kind of assumption — exemption, zero-rating, non-taxability, or deductibility (B), and reasonable, non-negligent grounds for that assumption (C). A careless mistake fails C even where it is honest; deliberate under-declaration fails A and never reaches B.
3. The proviso defeats the bar where an assessment issued inside the window. If the Commissioner assessed the unpaid tax within the six years, time stops mattering — the assessed debt is recoverable like any other. The bar protects only the unassessed old period.
4. It must be raised. The source Act's annotation on this very passage records that in Triangle Ltd & Hippo Valley Estates v ZIMRA & 10 Ors (on appeal, 21-SC-082) the "appellants failed to take advantage of this point for assessments made outside the 6 years." A limitation defence that is never pleaded protects no one — section E returns to this.

For old-period work this is the single most important provision: it means a VAT exposure in a legacy period more than six years stale may be legally dead — but only if the taxpayer can prove the A-B-C triad, only if no assessment issued in time, and only if someone in the room knows to say so.

B.4 The taxpayer's reach backwards: ITA Section 48 and VAT Section 44

The mirror-image clocks run against the taxpayer. Section 48 of the Income Tax Act (walked verbatim in the amendment lesson) compels a reduced assessment where overcharge is proved, but proviso (iii) caps the claim at 6 years, and proviso (ii) deems treatment that followed the practice generally prevailing to be proper — closing the door on practice-based reclaims. Section 48(3) adds 60-day refund interest (SI 212/2022; Delta 16-HH-378) with a carve-out where the taxpayer's own defective return caused the overcharge.

The VAT counterparts were confirmed verbatim this run. Section 44(1), proviso (a): no input-credit refund "unless the claim for the refund is made within 6 years after the end of the said tax period"; proviso (b): a de minimis amount (US$60 or the prescribed amount or less — the source text's ZWL figure prints ambiguously; ) is not refunded but carried forward (threshold history: increased FA 10/2020, FA 7/2021 Section 56, FA 8/2022). Section 44(2)–(3): overpaid tax, additional tax, penalty or interest is refundable on application, but only if the claim is received within 6 years after the date of payment — and, where the payment accorded with the practice generally prevailing at the time, within 6 months after that date. Recall also from the Taxpayer Account Management lesson: Section 44(7) withholds any refund while any return remains unfiled — which couples old-period refund work directly to the back-filing lesson.

B.5 Evidence and identity across the cutover

Three further confirmed rules frame the migrated archive:

• VAT Section 42 (confirmed verbatim this run): "The production of any document issued by the Commissioner purporting to be a copy of or an extract from any notice of assessment shall be conclusive evidence of the making of an assessment, and shall, except in the case of proceedings on appeal against the assessment, be conclusive evidence that the amount and all the particulars of such assessment … are correct." A migrated legacy assessment, produced by the Commissioner, proves itself. Your countervailing power is your own Section 37B/Section 57(3) archive — and the objection/appeal channel while it is still open.
• Part VIIIA, Sections 80C and 80I: electronic records are admissible with due weight (PIL 17-HH-213), and electronic lodgment/receipt timing rules date what happened when. The migrated record's evidential life rests on this platform.
• TIN continuity: the local guide confirms the TIN "replaces the legacy Business Partner Number." The taxpayer's legal identity is continuous across the rename — legacy-era obligations attach to the same person now keyed by TIN, exactly as Section 25D of the Income Tax Act and Section 26 of the VAT Act make liability independent of registration status (established in the deregistration lessons).

C.1 The three lanes of historical work — and which lesson owns each

A practitioner confronted with "a problem in an old period" must first classify it, because the cure, the channel and the cost differ completely:

1. Wrongly filed, SSP era → the amendment lanes (free in-window amendment; Section 48/VAT Section 44 downward; Section 47 upward; Case Management declarations; VDA01 for multi-period under-declaration). Owned by the Amending a Current Return lesson.
2. Never filed, SSP era → back-filing through Pending Tax Returns, with the five-gear consequence engine (Section 46(1)(a) / VAT Section 39(2)(a)(i), estimation, interest, the compliance grid, the aggravation curve). Owned by the Back-Filing lesson.
3. Pre-SSP period → old-period work. The period predates the portal; what exists in the SSP is whatever ZIMRA migrated in, visible on the Old Period Documents page. The legal machinery is the same statutes — but the procedural route differs because there may be no "pending return" line to click.

The classification question to ask is always: was the period inside or outside the SSP's life, and was a return filed for it? Note the trap inside lane 3: an unfiled legacy period is still an unfiled period. Section 37's duty to render returns, Section 45's estimation power and Section 46's additional tax do not expire just because the filing platform was retired; only the lattice of limitation rules (Section 47's six years absent misconduct; VAT Section 41's recoverability bar) closes old exposure — conditionally.

C.2 The six-year lattice: four clocks, four anchors

The recurring number in this lesson is six years, but it is four different six-year periods, and confusing their anchor dates is the classic old-period error:

Three structural observations:

• The clocks are deliberately asymmetric. The taxpayer's refund clock is hard (six years, sometimes six months, no misconduct extension in the taxpayer's favour), while the Commissioner's reassessment clock is soft (uncapped on misconduct). Parliament resolved doubt in the fisc's favour.
• The records clock should outlive your exposure, not match it. Because Section 47's bar lifts on alleged misconduct, the rational practice for any period where ZIMRA might allege non-disclosure is to retain records beyond six years — the statutory six is a floor backed by criminal sanction, not a destruction schedule. The defence to a fraud-era reassessment is evidence; destroying records at year six plus one day hands the Section 47(2)/Section 45 estimation machinery a free run.
• The anchors do not line up. A 2021 VAT period's records clock might expire in 2027 (last entry), its recoverability bar in 2027–28 (payability dates per period), its refund window at various 2027 dates (period-end / payment dates), and its reassessment exposure whenever six years from the relevant assessment lands — or never, on misconduct. Old-period analysis is therefore always provision-by-provision, date-by-date.

C.3 The migrated record: what it is and is not

The guide's sentence — historical returns "where ZIMRA has migrated them in" — defines the page's epistemic status. Treat the migrated archive as ZIMRA's system of record for the legacy era, with three consequences:

1. It is what the Commissioner will work from. An auditor reaching into a pre-SSP year starts from what TaRMS shows. If the migration is complete and accurate, that is your filed return; if a legacy return was filed but did not migrate, the system may show a gap that looks like non-filing. The cure for a false gap is evidence: the Section 37B archive, filing acknowledgments from the legacy platform, bank proof of payment.
2. It is not an admission by ZIMRA that balances are agreed. Migration is data transfer, not a settlement. A legacy-era debt that migrated into the Single Account remains recoverable (subject to the lattice); equally, a legacy credit that failed to migrate has not been forgiven — it has been misplaced, and Section 44/Section 48 windows for asserting it are still closing.
3. It interacts with the Single Account's oldest-first rule. As established in the introduction and Taxpayer Account Management lessons, payments allocate to the oldest debt first and are never directed. If migration carried in an old legacy balance — right or wrong — today's payments will feed it before they feed current liabilities, silently creating fresh arrears on current heads. This is the single most practical reason to reconcile the migrated record promptly: a disputed legacy balance distorts every payment you make until it is resolved.

C.4 The first-access reconciliation routine

Because of C.3, the first serious encounter with Old Period Documents should be a deliberate reconciliation, not a browse. A defensible routine:

1. Inventory the archive. List every migrated return by head, period and declared amounts.
2. Inventory your own record. Pull the Section 37B/Section 57(3) archive for the same periods: filed copies, acknowledgments, assessments received, proof of payments.
3. Three-way match per period: your filed return ↔ the migrated return ↔ the account balance trail (Summary Report and Tax Type Report in Taxpayer Accounting, which show assessments, payments and adjustments per head — established in the Taxpayer Account Management lesson).
4. Classify every mismatch into: migration gap (filed, not showing), balance error (payment not credited), genuine historic under-declaration (yours), or genuine historic over-payment (theirs to refund — check the Section 44/48 windows immediately, because they expire by date, not by discovery).
5. Route each class correctly: queries via E-Messaging; assessment disputes via Case Management within the 30-day objection window where a (re)assessment triggers it (Section 62; VAT Section 32); under-declarations via the disclosure doctrine (single period → amended declaration; multi-period → VDA01 — voluntariness is a wasting asset, as the back-filing lesson established); refund claims in writing on application (VAT Section 44(2): "on application by the person concerned").
6. Minute everything. The reconciliation file is itself Section 37B material — and the rebuttal evidence if conclusive-evidence documents (VAT Section 42) are produced against you later.

C.5 Currency eras inside old periods

Old periods straddle Zimbabwe's currency history — ZWL eras, the 2019–2020 transitions, USD dual-currency accounting under Section 37AA (established in Taxpayer Account Management), and the ZiG era. Two rules keep this navigable: currencies never net (the dual-ledger doctrine from the introduction lesson — a USD credit cannot settle a ZWL-era debt), and the law of the period governs the period — an old period is assessed under the rates, currency rules and thresholds in force then, not now. Old-period computations therefore begin by fixing the period's legal frame before touching any number. Where the era's figures cannot be confirmed from the source folder, flag rather than reconstruct from memory.

D.1 Individual: the landlord and the expiring refund

Tariro, a Harare landlord registered for income tax, discovers while browsing Old Period Documents that her migrated 2020 return shows tax assessed and paid of USD 4,800, but her own file shows the assessment was reduced on objection to USD 3,900 — the USD 900 difference was never refunded and never migrated as a credit.

• The claim is under Section 48 (overcharge proved by her objection outcome) and the refund machinery; the clock under the VAT-mirror logic of "claim within 6 years after the date of payment" (Section 48 proviso (iii) six-year cap) runs from her 2020 payment date.
• Payment was made, say, on 30 April 2020 → the window closes 30 April 2026. Discovered in June 2026, the claim is dead by lapse; discovered in March 2026, it survives — if lodged now, in writing, with the objection outcome attached.
• Lesson: old-period review is date-critical. The migrated archive did not destroy her right; her diary nearly did.

D.2 SME: the audit that reaches into the legacy era

Pamberi Hardware (Pvt) Ltd receives an audit notification covering 2019–2025. Periods to 2023 were filed on the legacy platform.

• The auditor works from the migrated record. Pamberi's defence file is its Section 37B archive — which, for its 2019 ledger (last entry February 2020), must by law still exist until February 2026. The audit lands in 2025: the records obligation and the audit overlap by design.
• For income tax, years more than six years behind their assessments are barred to ZIMRA unless misconduct is alleged (Section 47 proviso (ii)). If the auditor asserts wilful non-disclosure to reopen 2018, the battle is evidential — IAB 22-HH-032's warning that disclosure is "not split up into bits and pieces" cuts both ways: full contemporaneous disclosure in the legacy-era return is Pamberi's shield.
• For VAT, suppose the auditor finds zero-rating wrongly applied to USD 60,000 of 2018 supplies (output tax at the then rate, unassessed to date, payable more than six years ago). Section 41 bars recovery if Pamberi proves the A-B-C triad: no intent not to pay (A — the sales were invoiced and declared as zero-rated openly), good faith on an assumption of zero-rating (B — it relied on the exporter documentation regime), on reasonable, non-negligent grounds (C — it held the export documentation the practice required). If the triad holds and no assessment issued in time, the exposure is irrecoverable — but only if Pamberi raises Section 41. Triangle/Hippo Valley (21-SC-082) lost exactly this point by silence.

D.3 Large corporate: due diligence and the deceased-estate analogue

A purchaser's advisers performing due diligence on Mukonde Holdings treat Old Period Documents as the fastest map of the target's pre-SSP compliance: every migrated return is a declared position; every gap is a question. The buyer prices three findings: an unfiled legacy PAYE period (lane-3 exposure: Section 45 estimation + 13th Schedule para 10 tax-plus-equal-further-amount, as established in the PAYE lesson), a VAT over-claim within the Section 47/31 window (priced at tax + 100% Section 39 exposure + interest), and a stale 2017 exposure where Section 41's bar plausibly holds (priced near zero, with a warranty that the bar will be pleaded). The same archaeology serves an executor winding up a deceased trader's TIN (TIN-deregistration lesson): the final pre-death return must reconcile to the migrated history, and the executor's Section 56 personal-liability exposure makes the first-access reconciliation routine of C.4 non-optional before any distribution.

Honesty first: no reported Zimbabwean case deals with the Old Period Documents page or the TaRMS migration as such. The authorities are statutory-era and attach to the underlying provisions:

• Triangle Ltd & Hippo Valley Estates v ZIMRA & 10 Ors (on appeal, 21-SC-082) — annotated in the source Act directly against Section 41: the appellants "failed to take advantage of this point for assessments made outside the 6 years." The recoverability bar is a defence, not an automatic shield; counsel and practitioners must plead it. For old-period work this is the controlling practical authority. (Annotation-level only; .)
• NYS v ZIMRA 19-HH-617 — annotated against Section 37B: the record-keeping duty has teeth in litigation; the taxpayer who cannot produce the books from which returns were prepared fights uphill against estimation and additional tax.
• PIL (Pvt) Ltd v ZIMRA 17-HH-213 — annotated against both VAT Section 57 (records failure) and Section 80C (admissibility of electronic data): the evidential platform on which both ZIMRA's migrated record and the taxpayer's scanned archive stand.
• The Section 47 line ('T M Fee'; A v COT 00-ITC-1691; Deb 19-HH-664; Man 20-HH-078; Bath 20-HH-552; M Safaris 20-HH-331; IAB 22-HH-032; Zimplats 21-SC-159) — walked in the amendment lesson — governs when an old year stays closed and when misconduct reopens it indefinitely.
• Packers International 16-SC-028 (VAT Section 28(2)) remains the authority that a nil return is a return — equally true of legacy-era nils.

Where a proposition above rests on annotation in the source Acts rather than a read of the full judgment, treat it as a research pointer, not a substitute for the report.

1. Treating migration as agreement. "ZIMRA migrated my balance, so they accept it" — false in both directions. Migration is data transfer. Disputed legacy balances must be challenged through Case Management while objection windows live; missing credits must be claimed before Sections 44/48 close.
2. Destroying records on the sixth anniversary. Section 37B's six years is a criminal floor, not a destruction schedule. Because Section 47's bar lifts entirely on alleged fraud/misrepresentation/wilful non-disclosure, the only defence to a deep reassessment is evidence you chose to keep. Use Section 80H: scan, index, keep.
3. Failing to raise Section 41. The Triangle annotation is the cautionary tale — assessments for stale, unassessed VAT periods were not met with the recoverability bar, and the point was lost. Diarise the question for every old VAT exposure: did this become payable more than six years ago, was it ever assessed in time, and can we prove the A-B-C triad?
4. Confusing the four clocks' anchors. Counting reassessment time from the year-end (it runs from the assessment), or refund time from discovery (it runs from period-end/payment), or records time from filing (it runs from the last entry). Every old-period opinion should state the anchor date for each clock it relies on.
5. Letting the Single Account feed a disputed legacy debt. Oldest-first allocation means an unreconciled migrated balance silently consumes current payments, creating fresh arrears, ITF 263 failure and the Section 80 30% bleed on current trading. Reconcile first; pay into a clean ledger.
6. Assuming the practice-prevailing fuse is generous. Under VAT Section 44(3)(a)'s proviso, where the original payment followed the practice generally prevailing, the refund claim window collapses from six years to six months. Practice-based overpayments are nearly always unrecoverable by the time anyone notices; Section 48 proviso (ii) bars the income-tax analogue altogether.
7. Back-filing a legacy period blind. An unfiled pre-SSP period may already carry a Section 45/VAT Section 31 estimated assessment migrated into the ledger. As the back-filing lesson taught: an estimated assessment is displaced by objection (30 days, return as evidence), never by filing over it as if the estimate did not exist.

Work through the questions one at a time. Choose an answer and it is graded immediately, with an explanation and the provision it comes from. Your progress is saved, so you can stop and resume.

• The Old Period Documents page holds "historical returns from periods predating the SSP, where ZIMRA has migrated them in" — one confirmed sentence; everything screen-level beyond it is flagged for verification. The migration changed the platform, never the law.
• The six-year lattice has four clocks with four anchors: records (Section 37B / VAT Section 57(3) — last entry), reassessment (Section 47 — the assessment; uncapped on fraud/misrepresentation/wilful non-disclosure), VAT recoverability (Section 41 — date payable; A-B-C triad; defeated by an in-time assessment), refunds (Section 48 / VAT Section 44 — period-end or payment date; six-month practice fuse).
• Section 41 must be raised — Triangle/Hippo Valley 21-SC-082 is the standing warning that an unpleaded limitation protects no one.
• Section 37B's six years is a floor, not a destruction schedule; Section 80H makes permanent electronic retention cheap, and the only answer to an uncapped Section 47 reopening is evidence.
• The migrated record is ZIMRA's system of record and, via VAT Section 42, potentially conclusive — reconcile it against your own archive early, classify mismatches, route them correctly (E-Messaging / Case Management / written refund application / VDA01), and minute the exercise.
• Old-period analysis is provision-by-provision, date-by-date, currency-era by currency-era: fix the period's legal frame first; currencies never net; the law of the period governs the period.
• Three lanes, three lessons: wrongly filed (amend), never filed in the SSP era (back-file), pre-SSP (this lesson). Classify before you act.

The four six-year clocks compared

Old-period decision tree

flowchart TD
 A[Issue found in a historical period] --> B{Period inside the SSP era?}
 B -->|Yes, return filed wrongly| C[Amendment lanes - prior lesson]
 B -->|Yes, never filed| D[Back-filing via Pending Tax Returns - prior lesson]
 B -->|No - pre-SSP period| E[Old Period Documents: pull migrated record]
 E --> F{Migrated record matches your Section 37B archive?}
 F -->|Yes| G[Minute and close]
 F -->|No - credit or filing missing| H[Evidence pack; query via E-Messaging or Case Management; refund claim in writing - check Section 44 and Section 48 windows]
 F -->|No - under-declaration yours| I{Assessed by ZIMRA in time?}
 I -->|Yes| J[Debt stands; objection only if window open; settle or instalment]
 I -->|No| K{VAT amount payable more than 6 years ago AND A-B-C triad provable?}
 K -->|Yes| L[Raise Section 41 recoverability bar - Triangle warning: plead it]
 K -->|No| M[Disclosure doctrine: single period declaration or VDA01; pay with disclosure]

Statutes & sections

• Income Tax Act [Chapter 23:06] — Section 37B(1)–(2) duty to keep records, 6 years from last entry, penalty greater of level 7 / 10% of taxable income (confirmed verbatim; subst Act 3/2009; penalty subst Act 8/2011); Section 47 additional assessments (6-year bar; lifted on fraud/misrepresentation/wilful non-disclosure — walked verbatim in the amendment lesson); Section 48 reduced assessments/refunds (6-year claim; practice-prevailing bar; 60-day interest, SI 212/2022); Sections 45–46 estimation and additional tax; Section 37A(10)–(11) return-as-assessment; Section 25D liability independent of registration; Part VIIIA Sections 80C, 80H, 80I (electronic admissibility, retention, timing).
• VAT Act [Chapter 23:12] — Section 41 recoverability bar, conditions A–C and proviso (confirmed verbatim this lesson); Section 42 conclusive evidence of assessments (confirmed verbatim); Section 44(1)–(3) refunds, 6-year windows, US$60 de minimis carry-forward, 6-month practice fuse (confirmed verbatim; thresholds amended FA 10/2020, FA 7/2021, FA 8/2022); Section 44(7) refund withheld while any return unfiled; Section 57(3)–(4) records 6 years, alternative-form authorisation (confirmed verbatim); Sections 28–31, 32 returns, assessments, objections; Section 26 liability survives deregistration.

Case law

• Triangle Ltd & Hippo Valley Estates v ZIMRA & 10 Ors (on appeal, 21-SC-082) — Section 41 six-year bar lost where not raised (annotation in source Act).
• NYS v ZIMRA 19-HH-617 — Section 37B records duty (annotation).
• PIL (Pvt) Ltd v ZIMRA 17-HH-213 — VAT Section 57 records failure; Section 80C electronic admissibility (annotations).
• Section 47 case line and Packers International 16-SC-028 — as walked in the amendment and back-filing lessons.

ZIMRA guidance

• Comprehensive Guide to the ZIMRA Self-Service Portal (local source) — Tax Return Management five-page inventory; Old Period Documents description; TIN replaces legacy BPN. Official SSP online help (https://mytaxselfservice.zimra.co.zw/help/ssp/en/default.htm) unreachable this run — screen-level specifics flagged throughout.