Practice Management · Brief · Working level
Workpaper standards for credit claims you didn't build
What belongs in the CPA's file when a boutique performed the R&D or cost segregation study: the deliverable review, tie-outs to the return and the books, documented inquiries, and retention.
When the firm signs a return carrying a study it did not perform, the workpapers have one job: prove the reliance was reasonable. That is the standard Treas. Reg. §1.6694-1(e) sets for the Section 6694 defense, the standard Circular 230 §10.22 sets for due diligence, and — not coincidentally — the record a malpractice defense lawyer will ask for first. A file containing only the boutique's PDF and a tick mark proves the opposite.
The deliverable review
Get the complete study, not the summary letter, and memo a preparation-level review. The review is not a re-performance — the regulation is explicit that verification is not required — but it must engage the content:
- Methodology plausibility. For cost segregation, does the study reflect an engineering-based approach of the kind the Cost Segregation Audit Techniques Guide treats as most reliable, with asset-level detail and cited authority? For R&D, are business components identified and mapped to the four tests of Section 41(d), with named individuals and a visible basis for wage percentages?
- Consistency with what the firm knows. The preparer cannot ignore its own knowledge of the client. If the study describes systematic experimentation by a staff the firm knows spends its year on routine installations, that inconsistency must be raised, not filed.
- Sufficiency for the forms. Does the deliverable support the Section G business-component detail on Form 6765, or the class lives and conventions on Form 4562? A study the preparer must reverse-engineer to file from is itself a review finding.
Tie-outs
Reconciliations are where reliance becomes checkable, and they belong in the file as schedules, not assertions:
| Study figure | Tied to |
|---|---|
| Total QRE wages | W-2 Box 1 / payroll register for the named employees |
| Supply and contract research costs | General ledger detail; 65% limitation applied |
| Cost basis segregated | Fixed-asset ledger and closing statement |
| Credit computed | Form 6765 as filed, including the Section 280C election |
| Prior-year base amounts | Prior returns and prior studies |
Differences are not rounding trivia. A study whose wage total exceeds payroll is the "incorrect or incomplete" information that triggers the regulatory duty to inquire.
Questions asked and answered
Every review generates questions; the file should show them asked and answered in writing — email suffices. Typical entries: the basis for a 70% qualification percentage on a supervisor's wages; why funded-research contract terms were read as retaining rights and risk; how the study treated a building component the ledger shows was expensed. The Q&A trail is the single strongest artifact in the file, because it demonstrates the one thing the reliance regulation actually demands: that the preparer did not look away. If an answer is evasive or never comes, that too goes in the file — along with the resulting decision to adjust, disclose on Form 8275, or decline to sign. The disclosure decision memo closes the loop.
Alongside the return-year papers, cross-reference the firm's standing provider-vetting memo — the annual diligence described in our vetting playbook — so each year's file inherits the foundation without repeating it.
Retention
Retention must match the exposure's actual tail, which for credits is long. The assessment statute runs three years from filing, six for substantial omissions; the refund statute matters for amended claims; and a research credit carried forward under Section 39 keeps the original study operative for up to 20 years — an exam of a 2038 return can put a 2026 study on the table. Practical policy: retain all specialty-study files a minimum of seven years, and for credit studies, the life of the carryforward plus the open statute on the last year the credit is used. Store the deliverable in the format received, and log any provider updates or errata with dates. Circular 230 §10.28 and state board rules add client-record obligations on top — the IRS tax professionals hub collects the current guidance. Storage is cheap; reconstructing a defunct boutique's 2026 study in 2034 is not, and boutiques go defunct considerably more often than the credits they compute expire.
Frequently asked questions
- What workpapers should a CPA keep when relying on a specialist's credit study?
- The complete deliverable; a documented review memo covering methodology plausibility and consistency with what the firm knows about the client; tie-outs reconciling study totals to the return, payroll, and the fixed-asset ledger; the written questions asked of the provider and its answers; the disclosure decision; and the provider-vetting memo. Together these are the evidence that reliance on the specialist was reasonable.
- Does the CPA have to re-perform or verify the specialist's study?
- No. Treas. Reg. §1.6694-1(e) permits good-faith reliance without verification on information and advice furnished by another advisor. But the preparer may not ignore known facts that contradict the study and must inquire when information appears incorrect, inconsistent, or incomplete — so the file must show a real review and real inquiries, just not a re-performance.
- How long should specialty-tax workpapers be retained?
- At least as long as the exposure runs: the three-year assessment statute, extended to six years for substantial omissions, plus the 20-year life of a research credit carryforward under Section 39, which keeps the original study relevant to open years long after the claim year closes. Many firms set seven years as a floor and keep credit studies for the life of the carryforward plus the statute on the last year it is used.