AusIndustry does not review every R&D Tax Incentive registration. Registration is self-assessed and most applications are accepted on lodgement. AusIndustry then selects a subset for compliance activity, from a short information request to a formal examination and finding under the Industry Research and Development Act 1986.
Activity eligibility is decided by Industry Innovation and Science Australia (IISA) and its delegates, and a finding is binding on the ATO. Adverse findings can be challenged through internal review and then the Administrative Review Tribunal.
Who reviews R&D tax claims, and on what basis
Two agencies review different halves of the same claim. AusIndustry, acting for IISA, tests whether your registered activities meet the core and supporting activity definitions in s 355-25 and s 355-30 of the ITAA 1997. The ATO tests whether the expenditure you claimed against those activities is eligible and correctly calculated.
Neither agency approves your claim at registration. Receiving a registration number confirms your application was accepted, not that your activities have been assessed as eligible.
That distinction matters commercially. A company can hold a registration number, receive a refund, and still face an activity review two years later that unwinds both.
Two regulators, two questions
| AusIndustry (for IISA) | ATO | |
|---|---|---|
| Decides | Whether activities are eligible core or supporting R&D activities | Whether expenditure is notional R&D expenditure and the offset amount |
| Legal basis | Division 355 ITAA 1997 activity definitions; findings powers in the IR&D Act 1986 | Division 355 ITAA 1997 expenditure provisions |
| Typical trigger | Vague activity narratives, sector focus, repeat registrations of identical text | Expenditure spikes, associate payments, overseas amounts, feedstock and recoupment issues |
| Output | Finding on activity eligibility, binding on the ATO | Amended assessment, shortfall interest, penalties |
| Review path | Internal review, then Administrative Review Tribunal | Objection, then ART or Federal Court |
The two can run in parallel or sequentially. An ATO review that finds the expenditure is fine still leaves you exposed if AusIndustry later finds the underlying activity was not experimental. Our companion piece on ATO R&D tax incentive reviews covers the expenditure side.
Types of AusIndustry review
AusIndustry compliance activity sits on a spectrum. Understanding which type of contact you have received tells you how much is at stake and how quickly it can be resolved.
Registration completeness and pre-registration checks
The lightest form of contact happens before your registration is granted. Where a narrative field is blank, generic, or fails to address the question asked, AusIndustry may come back for clarification before processing the registration.
These checks are administrative rather than substantive. They are also a signal: a form that draws a completeness query has usually failed to hit the character minimums or answered a different question from the one asked.
The registration deadline is unaffected. Registration is due 10 months after the end of the income year, which is 30 April for a 30 June year end, and that date is statutory. See registration deadline dates.
Activity reviews and requests for information
The most common substantive contact is a request for information about one or more registered activities. AusIndustry names the activities in scope, asks how the company satisfied itself the outcome could not be known in advance, and requests supporting records.
This is where most claims are won or lost. Companies that can produce dated design documents, test logs and issue-tracker history usually close the review at this stage. Companies that respond with marketing material and reconstructed narratives usually escalate to the next stage.
Findings and examinations
The formal end of the spectrum is an examination leading to a finding under Part III of the Industry Research and Development Act 1986. A finding is a legally binding decision about whether registered activities were eligible R&D activities, and it is notified to the ATO.
During an examination, AusIndustry commonly refers the technical questions to expert assessors with relevant domain experience. That expert review step means your hypothesis and experiment are probed by someone who works in the field, so field-technical detail carries far more weight than commercial framing.
Findings are not only adverse. The same machinery produces advance findings (a binding decision obtained before registering, which binds for the income year of application plus the following two years) and overseas findings, which are mandatory before any offset can be claimed for R&D conducted outside Australia.
Both must be applied for before the end of the relevant income year, and no extensions are available. The overseas finding deadline is the single most commonly missed date in the program.
What triggers an AusIndustry review?
Selection is risk-based. AusIndustry does not publish a scoring model, but the patterns that draw attention are consistent and largely within your control.
Registration-level risk signals
- Activity descriptions that read as product roadmaps or feature lists rather than experiments.
- A hypothesis field that restates the project objective ("we wanted to build a faster engine") instead of naming a variable and a measurable threshold.
- Narrative text copied across multiple income years without change, or across multiple activities within one application.
- A single core activity covering an entire platform, rather than one discrete uncertainty per activity.
- Large or rapidly growing claims relative to the company's size and headcount.
- Registrations lodged at the deadline with the minimum characters typed into each field.
Sector and activity-level signals
Software attracts sustained attention because so much software work is competent application of known methods rather than experimentation. Activities that touch the s 355-25(2) exclusions also draw scrutiny: market research and market testing, management studies and efficiency surveys, activities associated with complying with statutory requirements or standards, reverse engineering an existing product, and internal administration software.
A structured study measuring a clinical or behavioural endpoint against a pre-stated hypothesis is not market research, even if it runs on A/B testing infrastructure. The same experiment measured on conversion, churn or revenue is excluded market research.
Similarly, being regulated does not itself trigger the standards exclusion. Activities undertaken to show compliance with a standard are excluded, while genuine experimentation that happens to occur in a regulated setting is not.
What AusIndustry asks for during a review
Reviewers work from your own registration text outward. They compare what you told IISA against what your records show actually happened, and any gap between the two drives the outcome.
The core activity fields under scrutiny
The current registration form, released 15 August 2025 in the customer portal, splits each core activity across separate fields with their own character minimums. Reviewers read them as separate assertions, not one narrative.
| Field | Form question | Minimum | What the reviewer tests |
|---|---|---|---|
| Sources investigated | Q79 | - | Was a worldwide prior-art search done before the work started, and what did it find |
| Why a competent professional could not know the outcome | Q80 | 600 | Is the uncertainty genuine against worldwide knowledge, not just new to your company |
| Hypothesis | Q82 | 650 | Is it specific, testable, falsifiable and measurable |
| Experiment | Q83 | 650 | Is a repeatable procedure described, with variables varied, held constant and measured |
| Evaluation | Q84 | 530 | Were results analysed against the hypothesis using established techniques |
| Conclusions | Q85 | 340 | What was concluded in this income period, including negative results |
| New knowledge | Q89 | 600 | New or improved material, product, device, process or service, distinguished from commercial value |
The evaluation field is the one tribunals most often find missing, as in GQHC and Commissioner of Taxation, where the observation and evaluation step of the systematic progression was central to the outcome.
Active Sports Management is the decision most often cited for scrutiny of whether the stated hypothesis was a hypothesis at all rather than a business objective.
Contemporaneous evidence the reviewer expects
Contemporaneous means created as the work happened, dated, and consistent with the narrative you filed. In Absolute Vision Technologies, discord between the registered description and the company's own records undermined the claim.
Records reviewers commonly ask for:
- Design documents or technical specifications recording the hypothesis before experimentation started.
- Prior-art search evidence: literature, patent searches, vendor datasheets, expert correspondence.
- Experiment records: test plans, results tables, benchmark output, trial protocols, clinical endpoint data.
- Development artefacts: commit history, pull requests, branch names, issue tracker tickets showing failed approaches.
- Meeting notes and decision records showing evaluation and conclusions.
- Time records apportioning staff hours to specific R&D activities.
- Contracts and invoices for contractor and associate work performed on the registered activities.
Rand builds the registration narrative from these artefacts as they are produced, so the story filed with AusIndustry is anchored to dated engineering and clinical evidence rather than reconstructed at year end. More detail sits in how to document R&D activities for AusIndustry and our guide to contemporaneous evidence.
How the review is assessed against the s 355-25 criteria
The reviewer applies the statutory test to each activity separately. Eligibility is self-assessed per activity, never per project or per product, so a review can accept two core activities and reject a third within the same registration.
The four elements assessed are:
- Experimental in nature. An experiment or set of related experiments, not routine development or application of a known solution.
- Outcome unknown in advance. Tested against a competent professional with access to knowledge publicly available or reasonably accessible anywhere in the world. New to the company is never enough.
- Systematic progression. Based on principles of established science, proceeding from hypothesis to experiment, observation and evaluation, and leading to logical conclusions. Ad hoc trial and error fails.
- Purpose of generating new knowledge. New or improved materials, products, devices, processes or services beyond what is publicly available.
Supporting activities are assessed separately and against a different test. Where the supporting activity falls in an excluded category or produces goods or services, it must have been undertaken for the dominant purpose of supporting the core activity. See core vs supporting activities.
Typical timeframes and what to expect
There is no legislated timeframe for an AusIndustry review. The practical range depends on the type of contact and how prepared the company is.
| Stage | Typical duration | What drives it |
|---|---|---|
| Registration completeness query | Days to weeks | Whether the missing field can be answered from existing records |
| Request for information on activities | 4 to 12 weeks | Number of activities in scope, quality of records supplied |
| Full examination toward a finding | Several months | Site visits or interviews, technical assessor involvement, iterative evidence requests |
| Internal review of an adverse finding | Weeks to months | Volume of new material submitted |
| ART external merits review | Months to over a year | Tribunal listing and complexity |
Expect a written request with a response date and the option to seek an extension. Reviews often also include a technical discussion where a domain assessor probes the hypothesis and experiment directly with the people who did the work.
Possible outcomes of a review
There are three practical outcomes: the review closes with no change, activities are partially accepted, or a finding is made that some or all registered activities were not eligible R&D activities.
An adverse finding is binding on the ATO. The Commissioner can then amend assessments to remove the notional deductions attributable to those activities, recover the offset already paid, and apply shortfall interest and penalties.
Worked example: what an adverse finding costs
A company with aggregated turnover under $20M and a 25% company tax rate claims $800,000 of notional R&D expenditure. At the refundable rate of 43.5% (company tax rate plus 18.5 percentage points), the offset is $348,000.
AusIndustry later finds that one core activity, carrying $300,000 of that expenditure, was routine development rather than experimental. The eligible base falls to $500,000 and the offset to $217,500.
The company must repay $130,500, plus shortfall interest, plus any penalty the ATO applies.
Review and appeal rights
An adverse finding is reviewable. The path is internal review by IISA or its delegate, then external merits review at the Administrative Review Tribunal, which replaced the Administrative Appeals Tribunal in October 2024. Deadlines for lodging apply, so act on the notice rather than the outcome you hope for.
Preparing before a review lands
The only reliable preparation is evidence created while the work is happening, matched to a registration narrative that answers each form question on its own terms. Both are cheaper to build during the year than to reconstruct under a request for information.
Rand assembles the registration from your engineering and project evidence as it accumulates, drafts each core activity field to its own statutory requirement, and has Expert Reviewers approve the application before it goes to AusIndustry. For the full picture of how registration fits the rest of the process, start with the R&D Tax Incentive explained end to end.
This article explains the rules and is not tax advice. Whether a specific activity qualifies depends on your facts, your records, and the state of worldwide knowledge in your field. Guidance on the program is published at business.gov.au and ato.gov.au.