TLDR: The AusIndustry R&D Tax Incentive registration is a structured online questionnaire in the customer portal at incentives.business.gov.au, organised as projects, then core R&D activities, then supporting activities. Each core activity is split across eight separate fields, from sources investigated through hypothesis, experiment and evaluation to new knowledge, several with character minimums between 340 and 650. Registration is due 10 months after the income year ends, 30 April for a 30 June year end.
What the registration form actually is
It is the application that registers your R&D activities with AusIndustry for a given income year. Activity eligibility is decided by Industry Innovation and Science Australia (IISA) and its delegates; the ATO separately administers expenditure and the offset. The current form version was released on 15 August 2025 and is completed entirely online in the customer portal at incentives.business.gov.au (see the R&D Tax Incentive on business.gov.au).
You cannot claim the offset without a registration number for the year. The registration comes first, always.
The form is not a free-text narrative. It is a questionnaire, and each question maps to a specific element of the statutory test in section 355-25 of the ITAA 1997.
Before you open the portal
Have these ready. Drafting inside the portal is the slowest possible way to complete the form.
- Your company details: ABN, ACN, incorporation status, financial year dates, contact people.
- A list of candidate activities, already split into core and supporting.
- The prior-art search results for each core activity: what you searched, where, and what you found.
- Contemporaneous records: design docs, test results, tickets, commit history, lab notebooks, trial protocols. See how to document R&D activities for AusIndustry for what these records should capture.
- Summary expenditure figures by category, and whether any payments went to a Registered Research Service Provider (RSP).
- Confirmation of whether any activity was conducted overseas (this needs an overseas finding, applied for before the end of the first income year in which the overseas work happens).
If you are unsure how to split your work into core and supporting, resolve that before drafting. See core vs supporting R&D activities.
Projects, core activities, supporting activities
The application has three nested levels, and understanding the hierarchy prevents most structural errors.
A project is a commercial container that groups activities. It is not an eligibility unit. Nothing about a project is assessed for eligibility.
A core R&D activity is the unit that is assessed. Eligibility is self-assessed per activity, never per project or per product. A typical application registers three to seven core activities.
A supporting R&D activity must be directly related to a named core activity. Where it falls in an excluded category, or produces goods or services, it must also satisfy a dominant purpose test.
The most common structural mistake is registering an entire product build as one core activity. One core activity should carry one discrete uncertainty.
Company and contact details
The first block of the application identifies the R&D entity and the people responsible for the claim. None of it is assessed for eligibility, but errors here delay processing and can invalidate the registration number your tax agent needs.
Expect to supply:
- The R&D entity's ABN, ACN and registered name, plus the income year being registered.
- The entity type you are claiming under, which confirms you are an eligible R&D entity (typically an Australian-incorporated company).
- Ultimate holding company details, where the R&D entity sits in a group.
- Registered tax agent details, if an agent is involved in the claim.
- A primary contact for the application and, usually, a secondary contact AusIndustry can reach about the activities.
Nominate a contact who can actually explain the technical work. Compliance questions land on this person.
Project-level fields
A project groups related activities so the application reads coherently. The fields here are descriptive and commercial in tone, which is the one place in the form where commercial framing is appropriate.
| Project field | What to put in it |
|---|---|
| Project name or title | The commercial or internal name of the body of work, not the uncertainty |
| Start and end dates | The period the project ran or is expected to run, which must sit consistently with the activities under it |
| Project objectives or description | What the project set out to achieve commercially, in plain language |
| Field of research and industry classification | The relevant ANZSRC research field and industry code for the work |
Keep the project description short. Every eligibility argument belongs in the core activity fields underneath it, not here.
What the form asks about each core activity
Each core activity repeats the same block of fields. Every narrative field accepts up to 4,000 characters, and several carry character minimums that each answer needs to satisfy.
| Field | Portal question | Character rule |
|---|---|---|
| Name of the core activity | Name | Maximum 250 |
| Describe the core R&D activity | Q63 | None stated |
| How did you determine the outcome could not be known in advance? | Checkbox | n/a (checkbox) |
| Sources investigated | Q79 | None stated |
| Why could a competent professional not have known the outcome? | Q80 | Minimum 600 |
| What is the hypothesis? | Q82 | Minimum 650 |
| What is the experiment and how did it test the hypothesis? | Q83 | Minimum 650 |
| How did you evaluate the results? | Q84 | Minimum 530 |
| Conclusions reached in the income period | Q85 | Minimum 340 |
| What evidence did the company keep? | Checkbox | n/a (checkbox) |
| What new knowledge was or will be produced? | Q89 | Minimum 600 |
That is roughly 3,400 characters of mandatory content per core activity, before the description and sources fields. Five core activities is a substantial drafting exercise.
Sources investigated and why the outcome was unknown
Two fields, two different jobs. The sources field (Q79) records the worldwide search you conducted before the activity: literature, patents, internet sources, expert input, and any domain-specific databases or datasheets. State what you found and why your specific question remained unanswered.
The competent professional field (Q80, minimum 600 characters) then applies the statutory test. Section 355-25(1)(a) asks whether the outcome could be known on the basis of current knowledge, information or experience. In practice this is tested against a competent professional with access to knowledge that is publicly available or reasonably accessible anywhere in the world.
"New to us" never passes. If the relevant knowledge sits inside a competitor as a trade secret and is not reasonably accessible, say so explicitly, because that is a legitimate basis for the gap.
A preceding checkbox asks how you determined the outcome was unknown: no information in literature or patents, experts advised no solution existed, no adaptable solutions, other, or did not look into existing knowledge. The last option undermines the rest of your answer.
Hypothesis, experiment and evaluation
These three fields carry the heaviest minimums and the heaviest scrutiny. The hypothesis field (Q82, minimum 650) must contain a testable, falsifiable proposition with a measurable target under defined conditions, not a restatement of the project objective.
A workable structure: it is hypothesised that [specific mechanism] will achieve [measurable target] under [defined conditions], where existing worldwide knowledge cannot predict whether this is achievable; the hypothesis is disproven if [failure condition].
Avoid "feasible", "viable", "scalable" and "cost-effective" as the core claim. Those are commercial framings, and a business objective dressed as a hypothesis is a recurring weakness in tribunal decisions. The Active Sports line of reasoning turns on exactly this point.
The experiment field (Q83, minimum 650) describes the method, not the outcome. Name the variables you varied, held constant and measured, show the progression across iterations, and include the attempts that failed. Avoid "trial and error" framing.
The evaluation field (Q84, minimum 530) is the one tribunals most often find missing, as in GQHC. Describe how results were analysed against the hypothesis using established techniques.
Conclusions and new knowledge
The conclusions field (Q85, minimum 340) asks what you concluded in this income period. Negative and partial results are fully eligible; a disproven hypothesis is a valid conclusion and often spawns the next one. If the activity is still running, say what has been concluded so far.
The new knowledge field (Q89, minimum 600) asks what new or improved material, product, device, process or service the activity produced, per section 355-25(1)(b). Keep it distinct from commercial value. "We shipped a feature customers liked" is not new knowledge; "we established the parameter range at which the approach holds latency under target" is.
For deeper drafting guidance across these fields, see how to write an R&D tax incentive technical narrative.
Supporting activity fields
Supporting activities have a shorter block. The form asks for a name (200 character maximum), a description of what was done and the main steps, a multi-select of which core activities it supports, and an explanation of how it directly supported them.
That explanation must name the part of the systematic progression it fed: the hypothesis, the experiment, the observation and evaluation, or the conclusions. "It was part of the same project" is not a relationship.
A dominant purpose field appears only when the activity falls in an excluded category under section 355-25(2) or produces (or directly relates to producing) goods or services. Where it appears, explain why supporting the core activity was the ruling, prevailing or most influential purpose, with cost apportionment where production is involved.
Character minimums, and why they exist
The minimums reflect the structure of section 355-25. A 650-character hypothesis field cannot be satisfied by a single sentence, which is precisely the point.
Two practical constraints:
- The form accepts plain keyboard characters only. Smart quotes, dashes and special symbols pasted from a word processor will be rejected or mangled. Draft in plain text.
- Padding to hit a minimum is visible. A reviewer reading a 700-character hypothesis that says nothing measurable will treat the whole activity as weak.
If you cannot write 650 defensible characters about the hypothesis, the honest conclusion is usually that the activity is not a core activity.
Expenditure questions on the registration
The registration collects expenditure information at a summary level only, and asks whether any payments were made to a Registered Research Service Provider or as a monetary contribution under the Cooperative Research Centre programme. Those two categories matter because they bypass the $20,000 minimum notional R&D spend that otherwise applies.
The detailed claim and the offset calculation are lodged separately with the ATO in the R&D tax schedule by your tax agent.
For context only: the offset itself (43.5% refundable for a base-rate entity with aggregated turnover under $20M, or the company tax rate plus a two-tier intensity premium at $20M and above) is calculated in that schedule, not in the registration. See how to calculate your R&D tax incentive rebate.
Sign-off and the declaration
The application is submitted by an authorised person for the company, who declares that the information is true and correct and that the registered activities meet the eligibility requirements. This is a self-assessment regime: AusIndustry does not pre-approve activities at registration.
That declaration is the reason the narratives must align with your records. In Absolute Vision, the tribunal focused on discord between the registered description and the contemporaneous evidence on file. Write what the records show, not what you wish they showed.
Common mistakes, field by field
- Activity name: naming a product or a sprint rather than the uncertainty being tested.
- Sources investigated: describing the search as internal only, or conducting it after the work.
- Competent professional: asserting difficulty rather than explaining why the outcome was unknowable.
- Hypothesis: a project goal with no measurable threshold and no failure condition.
- Experiment: describing the build, not the method, with no controlled variables.
- Evaluation: skipping straight from experiment to result with no analysis step.
- Conclusions: hiding negative outcomes, which are eligible and often the most credible content.
- New knowledge: describing commercial benefit instead of technical or clinical knowledge.
- Supporting activities: listing overheads with no named parent core activity.
For what to keep on file behind each of these, see what evidence AusIndustry accepts.
After you submit
You receive a registration number, which your tax agent needs for the R&D tax schedule. Registrations can be varied through the portal within the programme's limits and timeframes; a material change to activities or expenditure may also require an amendment to the lodged company tax return.
Two finding deadlines sit outside the registration process entirely and cannot be cured later:
| Finding | Apply by |
|---|---|
| Advance finding (binding eligibility decision, covers the application year plus the following two income years) | Before the end of the income year in which the activities are conducted |
| Overseas finding (mandatory for any overseas R&D) | Before the end of the first income year in which the overseas activities are conducted |
No late applications or extensions are accepted for either. The overseas finding is the single most commonly missed deadline in the programme.
If AusIndustry later reviews the registration and reaches an adverse eligibility decision, the path is internal review, then external merits review at the Administrative Review Tribunal, which replaced the AAT in October 2024.
Preparing the answers before you start
The form rewards preparation and punishes improvisation. Draft each field to its own minimum, against your own records, well before 30 April. See registration deadlines for the full calendar, and the complete R&D Tax Incentive guide for how registration fits the rest of the process.
Rand builds these fields directly from engineering and clinical evidence, maps each one to its portal question and minimum, and routes the draft to Expert Reviewers before the registration is lodged, so the narrative and the records say the same thing.
