Legal Analysis

How to Write an IRAC

9 min read
Quick Answer

How to Write an IRAC

An IRAC is a four-part unit of legal analysis — Issue, Rule, Application, Conclusion — that you repeat once for every disputed legal question in a problem. The Issue names the precise question, the Rule states the governing law at the level of generality the dispute requires, the Application matches specific facts to specific rule language, and the Conclusion commits to an answer.

What does each letter actually require?

Issue: a one-sentence question that names the legal test and the operative facts. Not "Is there a contract?" but "Whether an enforceable contract formed when Zehmer signed a bar napkin agreeing to sell his farm while both parties had been drinking." The facts in the issue statement tell the reader which sub-rule you are about to apply.

Rule: the legal standard, stated in the words a court would use, with authority attached. This includes the black-letter rule plus whatever gloss the cases add — definitions of terms, factor lists, burdens, and exceptions that the facts put in play.

Application: the reasoning that connects the rule's operative language to the client's facts, in both directions. This is the graded part. If your rule says "substantial certainty," the application must argue about substantial certainty using facts, not repeat the phrase.

Conclusion: a direct answer to the issue. "A court will likely find X" or "Dailey is liable for battery." Close questions still get an answer; hedging with "it could go either way" and stopping there forfeits points.

How do you frame the issue when there are five of them?

Do not write one giant IRAC per call of the question. Break the claim into elements, identify which elements are genuinely contested on these facts, and run a separate IRAC for each contested element. A battery problem has an intent element, a contact element, and a harm-or-offense element. If the contact is obvious and the intent is not, you write two sentences disposing of contact and a full IRAC on intent.

This is the single biggest structural decision you make on an exam. Professors seed facts that make one element close; those seeded facts are the scoring opportunities. Spending equal ink on every element signals that you cannot tell which is which.

Nest your IRACs to match the structure of the law. For a Fourth Amendment problem you might run an IRAC on whether there was a search at all, then, only if yes, an IRAC on whether a warrant exception applied, then an IRAC on the scope of the search. Signal the nesting with headings or transitions like "Assuming the stop was lawful, the next question is whether the frisk exceeded its permissible scope."

How do you state the rule without dumping your outline?

State the rule at the level of generality the dispute requires, then stop. If the fight is over the intent element of battery, you do not need the full history of trespass to the case. You need: a battery is an intentional harmful or offensive contact with the plaintiff's person, Restatement (Second) of Torts § 13; and intent is satisfied by purpose or by knowledge with substantial certainty that the contact will result, Garratt v. Dailey, 279 P.2d 1091 (Wash. 1955).

Synthesize rather than list. If three cases apply the same standard, extract the common principle and state it as one rule, then use the individual cases as illustrations in the application. A string of case summaries in the rule section is a signal that you have not decided what the law is.

Add rule explanation when the standard is fuzzy. For open-ended tests — reasonableness, minimum contacts under International Shoe Co. v. Washington, 326 U.S. 310 (1945), unconscionability — one or two sentences describing what facts a court found sufficient in a prior case gives your application something to compare against. Without that comparison, your application has nothing to grip.

On citation form: in a legal writing memo, follow Bluepages B10 for cases, B12 for statutes, and use a short form after the first full cite. On a closed-book exam, most professors accept a bare case name in the text; check the instructions, because this genuinely varies.

How do you write an application that scores?

Use the rule's exact operative words and pair each with a fact plus a because. Weak: "Here, Dailey acted intentionally." Strong: "Dailey knew with substantial certainty that contact would result because he saw Garratt beginning to lower herself into the chair and moved it anyway, in the fraction of a second before she sat — knowledge of an imminent sitting is what makes the resulting contact substantially certain rather than merely foreseeable."

Argue both sides where the facts are actually ambiguous, and then resolve. The structure that works is: strongest argument for one party, strongest counterargument, then the reason one prevails. "Dailey will respond that he was five years old and moved the chair in an attempt to help. But Garratt turns on actual subjective knowledge, not age-adjusted reasonableness, and the trial court's finding on remand was that he did know."

Do not restate facts you have not connected to a rule word. Fact recitation with no legal operator attached earns nothing. Conversely, do not assert a legal conclusion with no fact attached. Every sentence in the application should contain both.

Flag missing facts rather than inventing them. "The record does not say whether Dailey saw Garratt begin to sit; if he did not, the substantial-certainty theory fails and the claim reduces to negligence." That sentence demonstrates you know which fact controls.

What does a complete IRAC look like on the page?

Issue: Whether Dailey's moving the chair constituted the intent required for battery when he knew Garratt was in the act of sitting.

Rule: A battery is an intentional harmful or offensive contact with the person of another. Restatement (Second) of Torts § 13. The intent element is met if the defendant acted for the purpose of causing the contact or knew with substantial certainty that the contact would result. Garratt v. Dailey, 279 P.2d 1091 (Wash. 1955). A defendant's minority does not defeat intent; the inquiry is into what the child actually knew, not what a reasonable adult would have known.

Application: Dailey plainly lacked purpose — the facts indicate he moved the chair to sit in it himself, not to cause Garratt to fall. The claim therefore rises or falls on substantial certainty. Dailey was standing beside Garratt as she began to lower herself, and he removed the only object she was descending toward. Where a defendant removes a chair from a person already committed to sitting, contact with the ground is not a risk but a near inevitability, which is the line the substantial-certainty standard draws. Dailey will argue that a five-year-old cannot form that appreciation and that he attempted to return the chair, which suggests he did not anticipate the fall. But the attempted return is equally consistent with recognizing the danger a moment too late, and Garratt makes the question one of actual knowledge, which the finder of fact may infer from Dailey's proximity and unobstructed view.

Conclusion: A court will likely find the intent element satisfied, and Dailey is liable for battery.

When should you use CREAC or a different format instead?

For predictive memos and briefs, use CREAC — Conclusion, Rule, Explanation, Application, Conclusion. Legal readers want the answer first; burying it behind an issue statement makes them work. The added E, rule explanation, is a discrete paragraph or two describing how courts have applied the standard in analogous cases before you reach your facts. Variants you will hear named — CRuPAC, TREAT, IRREAC — differ mainly in whether they break out rule explanation and whether the thesis leads.

For timed exams, the format your professor wants genuinely varies, and this is worth an email in week two. Some professors want the conclusion first because it makes grading faster. Some deduct for issue statements longer than a line. Some want you to identify the jurisdiction split and analyze under both. Ask, and look at any model answers the professor has released — those are far better evidence than any generic advice.

IRAC is a floor, not a ceiling. Once you can hit the four parts reliably, the higher-scoring moves are policy arguments where the doctrine runs out, express treatment of jurisdictional splits, and remedies analysis after liability. Use IRAC as scaffolding for those; do not let it cap what you write.

What mistakes cost the most points?

The recurring failures are predictable, and every one of them is a structural choice rather than a knowledge gap. Watch for these:

  • Rule dumping — reciting everything you know about the doctrine, including elements no fact puts in dispute.
  • Conclusory application — "Here, the contact was offensive because it was offensive," with no fact doing work.
  • Equal time for every element, which hides your judgment about which one is close.
  • Omitting the counterargument on a fact pattern that was obviously built to be ambiguous.
  • Refusing to conclude, or concluding with "a court could go either way" and nothing more.
  • Analyzing under a rule you never stated, so the grader cannot tell whether you know the standard.

Key Takeaways

  • Run one IRAC per disputed element, not one per claim, and dispose of uncontested elements in a sentence.
  • State the rule only at the level of generality the dispute requires, with authority attached.
  • The application is the graded section: every sentence should pair a rule word with a fact and a because.
  • Argue both sides where facts are genuinely ambiguous, then resolve — never stop at "it could go either way."
  • Use CREAC with the conclusion first for memos and briefs; on exams, ask your professor which format they want.
  • Flag facts you would need rather than inventing them; identifying the outcome-determinative fact earns credit.

Frequently Asked Questions

Should the conclusion come first or last?
In a predictive memo or brief, first — legal readers expect the answer up front, which is why CREAC has displaced IRAC in most legal writing courses. On exams it varies by professor, and some explicitly prefer a lead conclusion because it speeds grading. Ask early in the semester and check any released model answers.
How long should each part be?
There is no fixed ratio, but on a timed exam the application should be roughly half to two-thirds of the IRAC's length. The issue is one sentence, the rule is usually two to five, and the conclusion is one. If your rule section is longer than your application, you are writing an outline rather than an analysis.
Do I need full Bluebook citations on an exam?
Almost never. Most professors accept a bare case name in the text — "under Garratt" — and some do not require case names at all if you state the rule correctly. Full Bluebook form is expected in legal writing assignments, journal work, and anything a practitioner would read. Confirm with the exam instructions rather than guessing.
What do I do when there is a jurisdictional split?
State both rules in the rule section, identify which jurisdictions follow which, and run the application twice — once under each. Then say which approach the exam's jurisdiction follows if the facts tell you, or explain which the court is likely to adopt and why. Splits are almost always planted deliberately, so ignoring one is a large deduction.
Is it wrong to write "IRAC" headings in a memo?
Yes. Use substantive headings that state your conclusion on the issue — "The court likely has specific jurisdiction because the defendant's website targeted forum residents" — rather than labels like "Rule" and "Application." The structure should be visible from the writing itself. Labeled headings are fine as a drafting tool that you delete before submitting.
How do I get better at the application specifically?
Take a model answer or a published opinion, delete the reasoning, and try to rebuild it from the facts and the stated rule, then compare. Also try the reverse: underline every fact in your own practice answer that is not attached to a rule word, and either connect it or cut it. Application improves through repetition on graded feedback, not through reading more about IRAC.

Related Guides