How to answer a legal problem question with IRAC

The problem question is law school's native assessment: a page of invented facts, a person with a grievance, and the instruction "advise the parties". It does not test whether you know the law. It tests whether you can apply the law to facts, which is a different skill, and the structure called IRAC exists because that skill has a repeatable shape. This page explains the shape, its Australian variants, what each stage is actually for, and the sentence patterns that carry you through each one.

What the examiner is testing

Read any marking rubric for a problem question and the same word appears: application. The examiner already knows the law. What they cannot know until they read your answer is whether you can spot which legal questions the facts raise, select the rules that govern them, and reason from rule to fact to conclusion. An answer that recites everything you know about negligence scores poorly even when every sentence in it is correct, because reciting is not applying.

IRAC in one paragraph

IRAC stands for Issue, Rule, Application, Conclusion. You identify the legal question the facts raise (Issue), state the law that governs it with authority (Rule), reason about how that law operates on these specific facts (Application), and commit to an answer (Conclusion). Then you do it again for the next issue. That is the whole method. Everything else on this page is about doing each stage well.

The Australian variants: ILAC and MIRAT

Australian law schools teach the same method under different names. ILAC (Issue, Law, Application, Conclusion) is IRAC with one label changed. MIRAT, developed in Australian legal education, is the same spine with two honest additions: it begins with Material facts, forcing you to sort the facts that matter from the decoration, and it ends with a Tentative conclusion, acknowledging that real legal answers carry uncertainty. If your subject teaches a specific variant, use that one. The differences are labels; the discipline is identical.

Issue: frame it as a question the facts raise

A good issue is specific to the facts and phrased as a question: "Whether the supermarket owed Mei a duty of care as an entrant on its premises" is an issue. "Negligence" is a topic. Most problem questions contain several issues, and finding all of them is itself assessed; a common instruction to yourself is to walk through the elements of each plausible cause of action and ask whether the facts put that element genuinely in dispute. Elements nobody could contest can be dealt with in a sentence; the contested elements are your real issues.

Rule: state, synthesise, cite

State the governing rule accurately and attach authority to it: the statute section where legislation governs, the case where the rule is judge-made. Where legislation and case law both speak, deal with the statute first and use the cases to interpret it. Two disciplines mark out strong answers here:

• Synthesise, do not list. "In case A the court said X. In case B the court said Y." is a reading log. "The test requires X, as qualified in later authority to include Y" is a rule statement.
• State only as much rule as the issue needs. The rule stage is not the place to prove you did the reading. Every sentence of rule you state is a promise to apply it later.

Application: where the marks live

Application means connecting a rule element to a specific fact with reasoning in between, and the most useful mechanical test is the word "because": "The risk was foreseeable because staff knew spills occurred in that aisle and inspections had lapsed for an hour." If a sentence in your application section could survive being moved to a different problem question, it is not application; it is rule restatement wearing the wrong hat.

The second discipline is to argue both sides where the facts allow it. Problem questions are drafted with deliberate weak points, and examiners place them expecting you to see them: "Against this, the supermarket may argue that the spill occurred moments before the fall, and no reasonable system of inspection would have detected it. That argument is weaker here because ..." An answer that meets the counterargument and explains why one side prevails demonstrates exactly the judgment being assessed.

Conclusion: commit, with honest hedging

Answer the question you framed. "It is therefore likely that a court would find the duty was breached" commits to a position and honestly marks its strength. Two failure modes sit either side of it: the fence-sitting non-answer ("it depends on the circumstances"), and false certainty ("the supermarket is clearly liable"). The professional middle register, and the small set of hedging phrases that produce it, is covered in detail in Legal English register. Match the hedge to your application: if your reasoning was one-sided, "likely"; if it was genuinely close, "on balance".

One IRAC per issue, not per answer

The classic structural mistake is a single giant IRAC: all the issues, then all the law, then a swamp of application. Run the cycle once per issue, in a sensible order (usually the order of the elements, with threshold questions first), each cycle a few paragraphs. Headings for each issue are welcome in most subjects and make your structure visible to a tired marker at 11pm.

A worked skeleton

Invented facts: Mei slips on spilled juice in a supermarket aisle. Staff last inspected the aisle over an hour earlier. The skeleton of one issue, with placeholders where your subject's authorities go:

> Issue: whether the supermarket breached its duty of care by failing to detect and remove the spill.

>

> Rule: breach is assessed against what a reasonable person in the defendant's position would have done in response to a foreseeable risk, weighing the probability of harm, its likely seriousness, and the burden of precautions [authority]. In most Australian jurisdictions the civil liability legislation restates this calculus [provision].

>

> Application: the risk was foreseeable because spills in supermarket aisles are a known and recurring hazard. The burden of precaution was modest: a system of regular inspection is inexpensive relative to the risk of injury. Against this, the supermarket may argue [counterargument on the facts]. That argument is weaker here because [fact-specific reasoning].

>

> Conclusion: a court would likely find the failure to inspect for over an hour fell below the standard of reasonable care.

Notice what the skeleton does not contain: no doctrine the issue does not need, no facts without a job, and no sentence that could be copied into a different answer.

Common mistakes, named

1. Rule dumping: three pages of law, one paragraph of application. Invert the proportions.
2. Fact repetition without reasoning: restating the facts next to the rule and hoping the connection is obvious. The "because" test catches this.
3. Missing the counterargument the facts were drafted to invite.
4. The universal conclusion: "it will depend on how the court weighs the factors." Commit.
5. Issue blindness from doctrine-first reading: deciding the question is "about offer and acceptance" and never noticing the capacity issue in paragraph two. Walk the elements.

A self-check before you submit

1. Count your issues, then count your conclusions. The numbers should match.
2. Find your application paragraphs and underline every "because". Sparse underlining means sparse application.
3. Check each rule statement carries authority, and each authority earns its place by being applied.
4. Check every conclusion commits, with a hedge matched to the strength of your reasoning.
5. Read only your headings and conclusions. A marker doing the same should be able to reconstruct your entire answer.

If you would like feedback on a problem answer before you hand it in, Legal Writing Lab is a free feedback tool for international law students writing Australian legal English. It flags clarity and register issues in your draft, explains each one, and never writes your work for you. Your words stay yours.