The definition of core and supporting R&D activities is the same in every sector: an experiment whose outcome cannot be known in advance, run as a systematic progression of work to generate new knowledge. The department publishes sector guidance because the places where that experiment hides, and the mistakes claimants make, differ by industry. This page walks through the main sectors; the industry guides go deeper.
Where qualifying work sits by sector
| Sector | Usually core or supporting | Usually excluded | Watch for |
|---|---|---|---|
| Software | Experiments on algorithms, architectures and performance where the outcome was uncertain | Internal administration software, routine development, use of known techniques | The internal administration exclusion; whole-of-project registrations |
| Manufacturing and engineering | New materials, processes and products tested through trials and prototypes | Routine production, quality control, tooling with known outcomes | Feedstock adjustment on trial output sold |
| Mining and resources | New extraction, processing or monitoring techniques tested experimentally | Prospecting, exploring and drilling for minerals or petroleum | The exploration exclusion |
| Life sciences | Preclinical and clinical experiments, formulation, device testing | Routine release testing, marketing studies | Overseas trials need an Overseas Finding |
| Agriculture and food | Trials of varieties, inputs, processes and formulations with uncertain outcomes | Normal farming, taste-driven recipe changes | Whole-of-farm claims; feedstock on crops sold |
| Energy and cleantech | Materials, storage, generation and grid experiments | Installing proven systems | Clawback for government grants |
Software
Software is the sector the department and the ATO have written most about, after years of registrations that described whole development programs as single core activities. The department's software guide says a core activity in software is an experiment: the team did not know whether a particular technical approach would achieve a result, could not find out from existing knowledge, framed a hypothesis and tested it. Building features, integrating systems, migrating platforms and fixing bugs are development, not experiment, unless a specific technical unknown had to be resolved by testing.
The statutory exclusion matters here: software developed, modified or customised for the dominant purpose of use by the company or its connected entities for the internal administration of their business functions cannot be a core activity. An ERP customisation or an internal reporting tool is out; a product sold to customers, or software that is part of a product, is not caught by the exclusion, though it still has to be an experiment. Agile teams should register activities at the level of the technical unknowns, keep the backlog and commit history tagged to them, and end the activity when the unknown is resolved.
Manufacturing and engineering
Manufacturers hold more eligible work than they claim, because process trials happen on the floor and are not called research. A trial of a new alloy through an existing casting line, a fixture designed and tested to hold a tolerance, a parameter study to run a line at a new rate with a new material: each is a core activity if the outcome was uncertain and a hypothesis was tested. Building and testing a prototype is a supporting activity, or part of the core activity where the prototype is the experiment.
Two Australian features apply. Where trial production is sold, a feedstock adjustment adds an amount to assessable income that reduces the benefit on the feedstock inputs used. And once the process runs in production, further improvements with known outcomes are ordinary engineering, so the activity should end when the hypothesis is resolved.
Life sciences
Drug, biotech and medical device development sits comfortably within the definition, and the department's guidance treats clinical trials as core activities where they test a hypothesis about safety or efficacy. The practical issues are location and funding. Trials run overseas qualify only with an Overseas Finding, applied for before the end of the income year, and the overseas spend must stay below the Australian spend across the project. Many biotechs are loss-making with turnover under A$20 million, so the refundable offset is often their largest source of funding, and the ATO scrutinises refunds accordingly. Grants from Australian governments for the same work trigger clawback adjustments.
Other sectors
Mining and resources. Prospecting, exploring or drilling for minerals or petroleum is excluded from core activities, whatever the technique. Developing a new drilling, sensing or processing technique and testing it experimentally can qualify, and the exploration activity that uses it may be a supporting activity if the dominant purpose test is met.
Agriculture and food. Field trials of varieties, feed regimes, irrigation and soil treatments are classic experiments, and the department's guidance warns against registering the whole farm or the whole season. Crops or livestock from trials that are sold attract the feedstock adjustment. Taste-driven recipe changes are excluded; formulation to achieve a technical result such as shelf life can qualify.
Energy and cleantech. Development qualifies in the ordinary way. Projects commonly carry ARENA or state grants, so the clawback adjustment should be modelled from the start.
Construction and engineering services. Novel structural or materials solutions can qualify, but the company that bears the risk and owns the results claims; work done to a client's specification and paid for regardless of outcome is generally the client's R&D or no one's.
Illustrative example
Illustrative example. Three hypothetical companies. Yarra Logistics builds a route-optimisation engine and does not know whether a constraint-programming approach can produce schedules within 30 seconds for 5,000 stops; it tests three formulations against a benchmark. Core activity, with the benchmarks as evidence; the booking interface built around it is development. Hunter Alloys trials a new heat-treatment schedule for a high-strength alloy, scrapping 60 test parts before hardness was consistent. Core activity; the scrap is consumed feedstock with no revenue, so no adjustment. Gippsland Dairy reformulates a yoghurt to taste sweeter. Excluded. Its parallel trial to extend shelf life to 60 days without preservatives, with microbial testing at intervals, is a core activity.
Related pages
Official sources
Educational content only, not tax advice. See the disclaimer.