.png)
IRAC is the sentence structure of legal analysis: Issue, Rule, Application, Conclusion. Every law school exam answer, whatever the course, is a chain of IRAC units, one per issue, and the format exists because it mirrors how graders award points. There is a point for framing the issue, points for the rule stated precisely, the largest pool for applying the rule to these facts, and a point for a committed conclusion.
Students do not lose exams because they lack a format. They lose them because they execute one or two of the four letters badly, almost always the A. This guide takes the letters one at a time, then shows a full unit built correctly.
A weak issue statement names a doctrine: whether there was a contract. A strong one names the collision between a specific rule and specific facts: whether the store's advertisement was a definite offer that Chen could accept by arriving first.
The second version does three jobs at once, it tells the grader which rule is coming, which facts matter, and that you saw why this fact pattern is not obvious. On a timed exam, a one sentence issue frame is enough; the finding of issues in the first place is its own skill, covered in issue spotting.
The rule section is a precise statement of the governing law, elements broken out, exceptions attached if the facts will trigger them. Two failure modes dominate. The rule dump, reciting every doctrine in the neighborhood to prove you studied, which costs time and signals that you cannot select.
And the case story, narrating what happened in the precedent, which earns nothing; cases appear in exam answers as one clause attached to the rule they stand for, not as plots. If a rule has three elements and only one is disputed, say so: elements one and two are plainly met on these facts, and the fight is element three. That sentence is triage, and graders reward it.
Application is where the majority of points sit, and the discipline is mechanical: every sentence in this section should contain a fact from the pattern. The reliable habit is the word because followed by a fact, the contact element is met because Smith grabbed Jones by the arm.
When the facts cut both ways, and on the issues the professor cares about they were built to, argue both directions before choosing: the strongest version of each side, each anchored in facts, then a reasoned pick. A paragraph of application that never touches the facts is a second rule section, and it scores like one.
One sentence. Take a position, flag genuine uncertainty honestly, a court would likely find, and move to the next issue. Graders rarely care which way you conclude on a close call; they care that you conclude, because refusing to signals that the analysis did not finish.
Facts: a store advertises one designer coat, worth eight hundred dollars, for one dollar, first come first served, and Chen arrives first; the store refuses to sell, saying advertisements are not offers.
The answer: The issue is whether the advertisement was an offer Chen accepted by arriving first, or a mere invitation to negotiate. Advertisements are generally invitations to deal, not offers, but an advertisement that is clear, definite, and explicit and leaves nothing open for negotiation constitutes an offer accepted by performance of the requested act. Here the advertisement was definite in every term that matters: a stated quantity, one coat, a stated price, one dollar, and a stated manner of acceptance, first come first served, so nothing remained to negotiate, and Chen performed the exact act requested by arriving first.
The store will argue the general rule and the absurd price signal that no reasonable person would take the ad literally, but the specificity cuts the other way, a named quantity and acceptance procedure is precisely what separates offers from puffery. A court would likely find an offer and acceptance, forming a contract at the advertised price.
Count the moves: issue as collision, general rule plus its exception, every application sentence carrying a fact, the counterargument given its strongest form, a committed close. Five sentences of analysis, full credit. Multiply by every issue in the pattern and that is the entire exam.
A three hour exam is fifteen to twenty five of these units. Run one full IRAC per issue, in the order the grading likely weights them, with headers so the grader can find each unit; the time allocation math is in exam strategy. When a single element needs its own analysis, minimum contacts inside personal jurisdiction, run a mini IRAC nested inside the larger one.
And when an issue is present but undisputed, spend one sentence, not a unit: battery is straightforward here because the blow was intentional and harmful, one line, move on. Knowing which issues deserve a unit and which deserve a sentence is most of exam judgment.
They mean they hate robotic IRAC: mechanical headers reading Issue and Rule, boilerplate transitions, units run on non issues. They still award points for the same four things in the same order, because the structure is the structure of legal reasoning, not a formatting preference.
Keep the skeleton, drop the labels, write in flowing paragraphs, and the answer will read as exactly what they asked for. The writing course version of this skeleton, built for memos rather than exams, is CREAC, and the differences are mapped in CREAC vs IRAC.
Short issue headers, yes, unless your professor's past exams or instructions say otherwise. Headers let a grader moving fast find and award your points, and they force you to organize by issue rather than by chronology.
Issue and conclusion, one sentence each. Rule, one to three sentences.
Application, as long as the disputed facts justify, typically two thirds of the unit. If your rule sections outweigh your application sections, the answer is inverted.
Not directly; policy questions want a thesis defended with arguments and counterarguments rather than rule application. The both sides discipline transfers; the skeleton does not.
Written by Ali, Georgetown Law, founder of Lovare Institut.
July 30, 2026
July 30, 2026