The R&D Tax Incentive for agtech and agriculture

The R&D Tax Incentive is one of the biggest levers the Australian government uses to encourage innovation, and it has a direct line to the paddock and the grow house. If your business is running experiments on-farm, breeding a new pasture variety, building a sensor network that predicts frost, or trialling a robotic weeder, you may be performing activities the ATO and AusIndustry define as eligible research and development. But getting the claim right asks for more than a good story. It wants structured evidence, a defensible link between cost and activity, and a registered tax agent who reviews and lodges on your behalf.

This guide steps through the process specifically for agtech and agriculture businesses. It is general information only, not tax, financial or legal advice. Eligibility rules, rates and thresholds can change, and every business is different. You must confirm your position for the current income year with a registered tax agent who can review your facts and lodge the claim. Nothing here guarantees an outcome or an audit-proof result.

Prerequisites

Before you walk through the steps, make sure you have these basics in place.

• An active Australian business (company structure is the most common vehicle, but other structures may also be eligible, check with your advisor).
• Cloud accounting that captures your R&D-related expenses. Xero, MYOB or QuickBooks are ideal because they let you, or a tool like GrantsMAX, isolate the costs linked to experimental activities.
• A registered tax agent or accountant who is comfortable with the R&D Tax Incentive and will review and lodge the claim. You own the claim; the agent reviews and lodges. If you do not already have one, GrantsMAX’s accountant channel can connect you with practitioners who work with agtech and primary producers.
• A habit of keeping contemporaneous records: trial notebooks, drone imagery, yield maps, lab reports, purchase invoices, timesheets. If you do not already keep these, start now. Retrospective evidence is much harder to build.

Now, the step-by-step walkthrough.

Step 1: Understand the R&D Tax Incentive basics

The R&D Tax Incentive is a self-assessment program jointly administered by the ATO and AusIndustry (part of the Department of Industry, Science and Resources). It is not a grant, it is a tax offset that reduces your income tax liability, and for eligible companies with aggregated turnover below $20 million (or below $50 million under the proposed 2026 reform, which is not yet enacted, refer to ato.gov.au for the latest), the offset is refundable. That means you may receive a cash refund if the offset exceeds your tax payable.

The program gives you a tax offset equal to a percentage of your eligible R&D expenditure, on top of the usual deduction. Put simply, if you spend dollar one on eligible R&D, you get to deduct that dollar like any business expense, and the incentive then gives you a further tax benefit. The exact rates are published on ato.gov.au and are adjusted from time to time; always verify them for the income year you are claiming.

To access the offset, you must register your R&D activities with AusIndustry within 10 months of the end of the income year, and then your registered tax agent lodges the claim with your company tax return. This is not a process you do alone. An agent must sign off, and the business owns the claim. Tools like GrantsMAX for first-time claimants can prepare the evidence-backed pack from your accounting data, but the review and lodgement rest with the agent.

Around the world, other countries run similar incentives. The United States, for example, has a federal research credit administered by the IRS (see IRS: Research Credit for the official overview). The legislative definition of qualified research is found in eCFR: Qualified research defined. While the Australian regime is different, the underlying idea, that systematic experimentation in a field of science or technology can earn a tax benefit, is similar. In agriculture, the US government also directly funds research through initiatives like the USDA NIFA: Agriculture and Food Research Initiative (AFRI). That underscores a global recognition that on-farm innovation deserves support.

Step 2: Identify core and supporting R&D activities on-farm

The legislation distinguishes between “core R&D activities” and “supporting R&D activities.” Core activities are the systematic, experimental work where the outcome cannot be known in advance and the purpose is to generate new knowledge. Supporting activities are those directly related to the core activities, for instance, taking measurements, maintaining equipment used solely for the trial, or collecting data that feeds the experiment.

Agtech and agriculture throw up dozens of examples that may fit. Consider:

• Trialling a novel cover-crop blend to determine whether it suppresses a particular weed species more effectively than the standard mix, where the underlying soil biology or allelopathic interaction is not well understood.
• Developing a predictive algorithm that uses on-farm weather, soil-moisture and plant-health data to optimise irrigation scheduling, provided the work follows a hypothesis and tests it.
• Breeding a new barley line with improved heat tolerance through marker-assisted selection, where the genetic outcome is uncertain and you are testing whether the trait truly expresses under local conditions.
• Building a prototype machine-vision system that identifies pest eggs on leaves in real time, where the technical feasibility is unknown.

These are the kinds of activities AusIndustry would examine against the “core R&D” definition. The supporting activities might include the daily data logging, lab analysis of leaf samples, calibration of sensors, and even the time spent by a farm hand exclusively tending the trial plot (provided that time is recorded and directly linked).

For many small and medium agribusinesses, the hardest part is separating the R&D from the routine farming. If you do the same thing in the same way every season, it is unlikely to be R&D. The moment you change a variable to test a hypothesis and measure what happens, you move into experimental territory. The GrantsMAX eligibility assessment reviews your accounting data against program rules and flags where a reviewer would look for a strong evidence trail, a useful early read before you involve your accountant.

Pro Tip: Write a one-paragraph hypothesis for each activity. For example: “We hypothesise that applying a seaweed extract biostimulant at the V3 growth stage will increase kernel weight in maize by at least 5% compared to an untreated control, by altering root-zone microbial activity.” If you cannot state what you are testing and why the answer is not already known, it may not be core R&D.

Step 3: Match activities to the legislative definition

Once you have a list, you check each activity against the formal definition. The law (Industry Innovation and Science Australia, informed by the Income Tax Assessment Act 1997) requires that core R&D activities:

• Are conducted for a purpose that involves a significant innovation or high level of technical risk.
• Involve a systematic progression of work based on principles of established science.
• Are carried out for the purpose of generating new knowledge (including new or improved materials, products, devices, processes or services).
• Have an outcome that cannot be known or determined in advance on the basis of current knowledge, information or experience.

In agtech, the “high level of technical risk” often sits in the biology, the environment or the interaction between hardware and a living system. A weather station that collects data is not R&D. But integrating that data into a model that predicts frost events with a novel statistical approach, where you genuinely do not know whether the model will work until you test it against historical and real-time observations, may well be.

The Department of Industry, Science and Resources publishes guidance on business.gov.au. AusIndustry also runs information sessions. You do not need a consultant to decide whether to proceed, but you do need a registered tax agent who can interpret the law against your facts. GrantsMAX’s discovery engine can help you surface which of your recorded activities might map to the R&D Tax Incentive, but the final call belongs to the agent.

Warning: Do not assume every agtech purchase qualifies. Buying an off-the-shelf drone and flying it over a crop to take photos is not R&D, even if the photos are useful. Using that drone imagery as input to a custom algorithm you are developing, where the algorithm’s performance is uncertain and you are systematically testing its accuracy against ground-truthed counts, is a very different picture.

Step 4: Document evidence that satisfies AusIndustry and the ATO

The phrase “evidence-backed” is not a buzzword, it is the backbone of a claim that can withstand review. The ATO has made it clear that contemporaneous records are expected. For on-farm R&D, this can include:

• Trial protocols and plot plans with dates, locations, and treatments.
• Observation logs, handwritten notebooks are perfectly acceptable if they are dated and legible.
• Weather records where relevant.
• Sensor output, drone imagery, and analysis reports.
• Lab test results from NATA-accredited facilities.
• Meeting notes where you discuss trial progress and decide on next steps.
• Invoices and receipts linked to specific R&D items (seed for the trial plot, not for the production paddock).
• Timesheets or diary notes that record who did what R&D work and for how long.

Supporting activities need their own documentation. If you rented a CO₂ sensor exclusively for a controlled-atmosphere storage trial, keep the rental agreement and a log showing the dates it was deployed. If a staff member splits time between routine work and R&D, use timesheets or diary notes to isolate R&D hours.

When GrantsMAX prepares a pack, it pulls the expense data directly from your cloud accounting (read-only) and layers in the narrative you provide. The result is a structured bundle that your registered tax agent reviews, refines and lodges through the Accountant Review & Lodge Workflow. The agent remains in control; the business owns the claim.

International practice reinforces why this matters. In the US, the IRS Research Credit also demands contemporaneous records, and the regulations at eCFR spell out the qualified research definition in detail. Agricultural extension services such as eXtension Foundation emphasise that recording experimental methods is as important as the results themselves. The same principle applies here: AusIndustry and the ATO will look for a clear narrative that links the money spent to the systematic experimentation.

Pro Tip: Use a cloud platform to store evidence as you go. A shared folder with subfolders by activity and date is simple and effective. If you are already on Microsoft 365 or Google Workspace, GrantsMAX can read metadata from those environments (read-only) to help corroborate your timeline.

Step 5: Register your R&D activities with AusIndustry

This step is non-negotiable and comes with a deadline: you must register your R&D activities with AusIndustry within 10 months of the end of your income year. For a 30 June year-end, that means by 30 April of the following year. Registration is done online through the business.gov.au portal, and you will need:

• Your ABN and company details.
• A description of each core and supporting activity you intend to claim.
• The income year to which the activities relate.

While the registration itself is simple, the description matters. It sets the scope of what you can claim. AusIndustry provides a registration guide, and you should use language that mirrors the legislative definition. Many businesses ask their registered tax agent to draft or review the registration text before hitting submit. If you use GrantsMAX for SMBs on cloud accounting, the platform can pre-populate activity descriptions based on your data, but your agent is still the person who should sign off before lodgement.

Do not miss the deadline. Late registrations are possible only in very limited circumstances and usually require a formal extension application. The cost of missing it is the loss of the offset for that year.

Step 6: Calculate the notional R&D deduction and tax offset

With activities registered, you tally the eligible expenditure. The main categories are:

• Employee costs: salary, wages and on-costs for the time staff spend directly on R&D.
• Contract expenditure: payments to third parties for R&D performed on your behalf, but be aware of the special rules that limit the amount you can claim if the contractor is an associate.
• Depreciation of assets used exclusively for R&D, or the decline in value of a tangible asset used partly for R&D (apportioned).
• Consumables and materials transformed or expended during the R&D.
• Overhead items like power and rent, but only if they are directly attributable to the R&D, this is rarely straightforward and should be discussed with your agent.

For agtech, the eligible expenses often include:

• Seed, fertiliser and chemicals used solely in trial plots.
• Prototype sensors or components you built to test a new monitoring technique.
• Subscription fees for a cloud platform you are using to run computational experiments (provided the platform is not a general business tool).
• Specialist consultant fees, such as an agronomist who designs and analyses a trial where the outcome is uncertain.

You must apportion any cost that serves a dual purpose. If a tractor is used 20% of the time for R&D plot work and 80% for production, only 20% of its decline in value can be claimed. Keep a logbook or GPS track log to substantiate the split.

The tax offset is then calculated as a percentage of the total notional deduction (the total R&D spend you are claiming). The percentages are set by legislation and differ depending on your entity type and aggregated turnover. As always, check ato.gov.au for the current rates.

If you are a small or medium agribusiness on Xero, MYOB or QuickBooks, the numbers are already in your ledgers. GrantsMAX reads that data, applies the apportionment logic you define, and builds a schedule that your agent can review. It does not lodge; it prepares.

Step 7: Prepare the claim with your registered tax agent

At this point you have:

• Registered activities.
• An evidence pack containing narratives, expense schedules, trial logs, and supporting documents.
• A clear delineation between R&D and routine work.

Now your registered tax agent reviews everything, may ask for more detail, and then lodges the claim as part of your company tax return. The agent is the one who signs the return; you remain the owner of the claim and are responsible for its accuracy.

For many agtech businesses, the claim is not huge, but it is material. A seed startup running three trial sites, a vineyard testing a new disease-monitoring AI, or a family farm experimenting with methane-reducing feed additives, each could have $50,000 to $200,000 of eligible R&D expenditure in a year. That translates into a meaningful offset that can fund the next round of experiments.

Accountants who work with primary producers often tell us the main barrier is the time it takes to pull the evidence together. GrantsMAX for founders and CFOs is designed to flatten that barrier: it connects to your accounting data, surfaces the activities that may be eligible, and drafts the narratives, so your agent receives a nearly-ready pack instead of a shoebox of receipts.

The agent lodges; you get the benefit. And because the Annual Refresh & Accountant Channel can roll the process forward each year, your record‑keeping habits compound into an ever-stronger evidence base.

Pro tips and common pitfalls

Pro Tip: Start small. If you have never claimed the R&D Tax Incentive, pick one well-documented trial and work through the steps with your agent. Once you are comfortable, you can add more activities.

Pro Tip: Involve your accountant early. Let them see the trial protocols and spending before the end of the financial year. They can tell you whether the records you are keeping are fit for purpose.

Warning: Do not double-count. If you receive a government grant for the same project (for example, a state innovation grant or an EMDG export grant), the R&D expenditure funded by that grant may be excluded from the R&D Tax Incentive calculation. Austrade’s EMDG guidelines and the R&D legislation interact; ask your agent to review any overlap.

Warning: The “dominant purpose” rule. If an activity has more than one purpose, you must show that the dominant purpose is generating new knowledge. If the real driver is producing a commercial crop and the experiment is just a side observation, AusIndustry may not accept it.

On-farm and agtech experimentation examples

To make all this concrete, here are a few real-world-style scenarios drawn from Australian agriculture.

Example 1: Soil carbon measurement trial. A mixed cropping and grazing operation wants to know whether a specific combination of cover crops and rotational grazing will lift soil organic carbon to a level that generates Australian Carbon Credit Units. They design a randomised block trial with control paddocks, baseline soil cores, and a hypothesis that the treatment will increase soil carbon by at least 0.2% over three years. The outcome is genuinely uncertain because local soil type, rainfall and microbial communities could confound the result. The core R&D activities are the systematic manipulation and measurement of variables; supporting activities include the lab analysis of soil cores and the detailed stock movement records needed to separate trial paddocks from production paddocks.

Example 2: In-field disease detection AI. A horticulture technology company develops a smartphone app that uses a convolutional neural network to detect early signs of powdery mildew in grapevines. The team captures thousands of leaf images under varied lighting, labels them with ground-truth PCR test results, and iterates on the model architecture. They record the training runs, the hyperparameters tested, and the performance against a holdout set. The core R&D lies in the systematic experimentation to build a model that can generalise; buying the developer’s laptop is not R&D, but the cloud compute hours consumed by the training runs are eligible as consumables or contract expenditure.

Example 3: Alternative protein from pulse waste. A food business mills chickpea processing byproduct and tests several enzymatic hydrolysis conditions to see whether it can produce a protein isolate with a neutral taste profile. The company documents each treatment combination, measures protein yield and sensory scores, and iterates based on the results. Because the outcome, whether the process works at pilot scale with an acceptable flavour, cannot be known in advance, the work may be core R&D. The raw material cost, the enzyme costs, and the labour of the food scientist are all eligible, provided they are directly linked to the experiment.

These examples show that the R&D Tax Incentive reaches far beyond the lab coat stereotype. Your paddock, packing shed or processing line can all be the site of eligible R&D.

Internationally, agricultural research is well supported. The USDA National Agricultural Library maintains an alternative farming systems information centre that catalogues innovation in production practices. The AgMRC provides practical guides on farm business and technology. Academic journals like the MDPI Agronomy Journal and research articles on ScienceDirect demonstrate the depth of global scientific inquiry in agriculture. The Australian system is different, but the principle is the same: government wants to encourage the experiments that push the industry forward.

Key takeaways and next steps

• The R&D Tax Incentive is a self-assessed tax offset, not a grant. You must register with AusIndustry and lodge through a registered tax agent.
• Eligibility hinges on systematic experimentation where the outcome is not known in advance. Agtech and on-farm trials often meet this test, but routine farming does not.
• Evidence is everything. Keep contemporaneous records, trials books, images, receipts, timesheets, and tie every dollar to a documented R&D activity.
• The claim process follows a clear order: register, gather evidence, calculate eligible expenditure, prepare the pack, have your agent review and lodge.
• The agent lodges; you own the claim. An AI tool like GrantsMAX can prepare the pack from your accounting and collaboration data, but it never lodges or guarantees an outcome.

If your agtech or agricultural business is running experiments and you want to know whether they may be eligible, the most productive next step is to talk to your registered tax agent. If your books live in Xero, MYOB or QuickBooks and you want to see what a data-ready claim looks like, you can join the GrantsMAX waitlist. The platform will connect to your data, surface what you may be eligible for, and hand your accountant a structured, evidence-backed pack for review and lodgement. Why GrantsMAX explains the difference from legacy consultants and grant directories.

Ready to turn your on-farm experiments into a substantiated claim? Join the GrantsMAX waitlist and let your data do the heavy lifting.