Quick answer: For an R&DTI claim you need contemporaneous records that prove two things — that your activities were eligible R&D (technical-uncertainty memos, experiment plans and results, project and test notes) and that the claimed expenditure relates to them (timesheets, general-ledger extracts, contractor invoices and a written apportionment method). Keep them in English for around five years.
25 June 2026 — the 2026–27 Federal Budget announced R&DTI changes scheduled to start 1 July 2028; this article describes the current rules unless stated otherwise.
The most expensive mistake companies make with the R&D Tax Incentive (R&DTI) is not getting the science wrong — it is leaving the paperwork until the accountant asks for it. By then the year is over, the engineers have moved on, and the story has to be reconstructed from memory. Under a self-assessment program, that reconstruction is exactly what does not hold up.
This is general information from a Registered Research Service Provider, not tax advice. The mechanics below come from the ATO, AusIndustry and business.gov.au, and your own eligibility always depends on your circumstances — you self-assess.
The bottom line: records carry the burden, not your registration
Registering your activities with AusIndustry is a procedural step. It is not a finding that your activities or expenditure are eligible. If the ATO or AusIndustry reviews your claim, the onus sits with you to substantiate it — and you substantiate it with records, not with the registration number.
The ATO is explicit that you must keep records that show your eligible R&D activities took place and that the expenditure you claimed actually relates to those activities. Two threads have to connect:
The activity thread — evidence that you were resolving a genuine technical uncertainty through a systematic, experimental process (a core R&D activity), and that supporting activities were directly related to it.
The expenditure thread — evidence linking the dollars you notionally deducted to those specific activities.
A claim falls over most often not because the science was weak, but because nobody can draw a credible line between the two.
What "contemporaneous" actually means
Contemporaneous records are records made as the work happens — not written up after year-end to justify a number. This is the single concept that separates a defensible claim from a fragile one.
The ATO's guidance, Keeping records and calculating your notional deductions, and business.gov.au's Records to show your R&D activities are eligible both point the same way: the documentation should be a by-product of doing the work, generated in real time. A timeline that shows a hypothesis recorded before the experiment, results captured during it, and conclusions drawn after is inherently more credible than a polished narrative dated the week before lodgement.
Practical requirements worth anchoring to:
Records should be kept in a form that is accessible and can be reasonably understood or translated if required.
You generally need to retain R&D records for at least five years from when they were prepared or obtained, and longer if a review or dispute is ongoing.
They must be specific enough to link activities to expenditure, not just describe the project in general terms.
The evidence checklist: what to keep while the work is live
There is no prescribed template, and you do not need a bespoke system. You need to be able to produce, for each registered activity, evidence across roughly these categories. The table below is a quick reference you can use to check your own record trail:
Evidence typeExamplesWho owns itWhere it's storedActivity evidenceTechnical-uncertainty memos, hypotheses, experiment plans, test results, project notesTechnical / engineering leadProject repo, lab notebook, design docs, issue trackerExpenditure evidenceTimesheets, GL / cost extracts, written apportionment method, contractor invoicesFinance / payrollAccounting system, payroll, GL, contract files
Activity evidence (proves the science)
Technical-uncertainty memos — a short written statement, dated, of the unknown you could not resolve from existing knowledge or competent professional practice.
Experiment plans and results — your hypothesis, what you tested, observations, and conclusions. Failed experiments are evidence too; the R&DTI does not require success.
Project notes and meeting records — design decisions, version history, commit logs, lab notebooks, test logs, prototype iterations.
Expenditure evidence (proves the dollars)
Timesheets or time-tracking apportioned to R&D activities — Questions such as “can I claim if I didn’t track hours?” often become difficult in practice, because you still need to demonstrate a reasonable basis for apportioning time and cost to R&D activities.
General-ledger and cost extracts identifying salaries, contractor invoices and overheads tied to the work.
Your apportionment methodology — a written, reasonable basis for splitting mixed costs (e.g., a developer who spent 40% of their time on eligible work) between R&D and business-as-usual. The method matters as much as the percentage.
If you can hand a reviewer that bundle and trace one activity to its expenditure, you are in a strong position. If a category is missing, that is where questions start.
What tends to trigger a review
The ATO does not publish a checklist of common review risk indicators, but its compliance commentary and the pattern of reviews point to recurring red flags: claims where activities are described in commercial language with no technical uncertainty; whole-of-business or whole-of-salary apportionment with no methodology; ordinary software, internal admin or excluded activities dressed up as R&D (see what does not qualify for the R&D Tax Incentive); and large year-on-year jumps unsupported by contemporaneous records. The common factor is almost always thin or retrospective documentation.
The Adelaide RSP angle: build the record while you build the science
This is where structuring the work early pays off. A Registered Research Service Provider like Ignition Research — based at Lot Fourteen in Adelaide and listed on the AusIndustry register as RSP000047 — helps frame the research itself: defining the technical uncertainty up front, designing the experimental approach, and setting up the record trail so the evidence exists as a by-product of the work, not as a year-end scramble. When the science and its documentation are designed together, the contemporaneous record largely writes itself.
For the broader picture of how the program works, our R&D Tax Incentive overview sets out eligibility, registration and timing. None of this replaces advice on your own tax position — you should self-assess and seek your own advice.
Talk to Ignition Research before you lodge — we help structure and evidence the research so your claim is defensible. This is research-side support from a Registered Research Service Provider, not tax advice; you self-assess your own eligibility.
Frequently asked questions
Q: What records do I need for an R&D Tax Incentive claim? A: Records that prove two things: that your activities were eligible R&D (technical-uncertainty memos, experiment plans and results, project and test notes) and that the expenditure you claimed relates to them (timesheets or time-tracking, general-ledger extracts, contractor invoices, and a written apportionment methodology). The ATO and business.gov.au expect the records to link activities to expenditure, in English, and kept for around five years.
Q: What are contemporaneous records? A: Records made as the work happens, not reconstructed after the financial year ends. A hypothesis logged before an experiment, results captured during it, and conclusions written after carry far more weight than a narrative drafted just before lodgement. Contemporaneous documentation is the strongest evidence in a review.
Q: What are common review risk indicators for an R&D claim? A: The ATO does not publish a fixed list, but reviews commonly follow thin or retrospective records, activities described without genuine technical uncertainty, whole-of-salary apportionment with no methodology, excluded or ordinary activities claimed as R&D, and unexplained year-on-year increases. Strong contemporaneous evidence is the best protection.
Q: How long must I keep R&D records? A: Generally around five years from when you prepared or obtained the records, and longer if a review or dispute is underway. Keep them in English and in a form you can produce on request. Confirm the current rule with the ATO for your situation.
Sources & further reading
Keeping records and calculating your notional deductions (ATO)
Records to show your R&D activities are eligible (business.gov.au)
This article is general information from a Registered Research Service Provider about the R&D Tax Incentive. It is not tax, legal or financial advice; eligibility depends on your circumstances and you should self-assess and seek your own advice.
