How to Research Case Law: A Step-by-Step Method

To research case law, frame the legal issue precisely, search for on-point decisions, read the headnote and holding of each promising case, expand to the authority it cites and that cites it, filter for binding decisions in your jurisdiction, and verify that every case you keep is still good law. Good research is a repeatable method, not a lucky search.

Anyone can type words into a search box. The difference between a quick lookup and real legal research is a disciplined process that turns a vague question into a short list of cases you can actually rely on. Here is a method you can run every time.

Step 1: Frame the issue and jurisdiction

Before you search, get specific about two things: the exact legal question and the jurisdiction whose law controls. "Is a non-compete enforceable" is too broad. "Is a non-compete enforceable against an employee who was laid off, in this particular state" is a researchable issue.

Jurisdiction matters just as much, because a case only binds courts below it in the same system. A perfectly on-point decision from another state may be persuasive at best. Nailing down the issue and jurisdiction first saves hours, and it draws directly on the ideas in binding vs persuasive precedent.

Step 2: Search for on-point cases

Now find candidate cases. You have three main routes, and skilled researchers mix them:

• Plain-English search. Ask the question the way you would say it out loud and read the precedents that come back with summaries and citations. This is the fastest way to map the landscape, and it is the approach behind plain-English case search in Cases.
• Keyword and boolean search. Build a precise query with terms and connectors when you know the vocabulary of the area.
• Topic and digest search. Once you have one good case, use its topic classification to pull related authority.

For a fuller comparison of free and paid ways to run this step, see how to find case law.

Step 3: Read the headnote, then the holding

You will get more results than you can read in full, so triage. Start with the headnote, the editorial summary at the top of a case, to decide whether a decision is even relevant. Headnotes are written by publishers, not courts, so they are a filter, not the law itself.

For every case that survives the filter, read the holding in the court's own words. The holding is the rule the case actually stands for. Distinguish it from dicta, the court's asides that are not necessary to the decision and do not bind. Confusing memorable dicta for the holding is one of the most common research errors. Tools that surface the summary and holding speed this up, but you still confirm the point against the opinion.

Step 4: Expand from your best case

A strong case is a hub. Two moves grow your research quickly:

• Look backward. Read the cases the opinion cites. These are the authorities the court relied on and are usually foundational to the issue.
• Look forward. Find the cases that cite your case. These show how later courts applied, extended, narrowed, or criticized it.

Looking forward does double duty. It surfaces newer authority, and it starts your good-law check, because a later case that criticizes or overrules your case is exactly what you need to know about.

Step 5: Filter for authority and relevance

Now weigh what you have found. Two cases can address your issue and still carry very different weight:

1. Binding beats persuasive. Prefer a decision from a higher court in your jurisdiction over an out-of-jurisdiction case, even a factually closer one.
2. Higher courts beat lower courts. A state supreme court decision outranks an intermediate appellate decision within the same system.
3. On-point facts beat general statements. A case with facts like yours is more useful than a case that mentions the issue in passing.
4. Recent beats stale, all else equal. A newer decision is more likely to reflect current law, though a landmark older case can still control.

This filtering is where research becomes analysis. You are not just collecting cases; you are ranking them by how much a court would have to respect them.

Step 6: Verify good law

The final step is non-negotiable: confirm every case you plan to rely on is still good law. A decision can be overruled, superseded by statute, or undercut by later cases. Traditionally you check this with a citator such as Shepard's or KeyCite, which flags negative treatment with warning signals. Free tools like Google Scholar and CourtListener also show citing references you can review. Our guide on how to Shepardize a case walks through the mechanics and the warning signals.

Verification also means confirming the case actually exists and says what you think it says. This has become critical because AI tools can generate convincing but fake citations, a problem we cover in AI hallucination in legal cases. Read the opinion, confirm the citation, and check the holding at the source. Tools with checkable citations that link straight to the source make this the work of a moment rather than a step people skip.

Putting the method together: a short example

Say the question is whether a landlord is liable for a slip-and-fall in a shared hallway, in your state. You would frame the issue and confirm the state, search in plain English for premises-liability cases about common areas, read headnotes to find the on-point ones, read each holding to see what duty the court imposed, follow citations backward and forward to find the leading decision and any recent developments, prefer your state's highest court, and finally verify the leading case is still good law. What you end up with is a small, ranked set of reliable authorities rather than a pile of tabs. (This is a general illustration of method, not advice about any real dispute.)

Common research mistakes and how to avoid them

Even experienced researchers fall into predictable traps. Watching for them makes your work more reliable:

• Stopping at the first good case. One on-point decision feels like victory, but a single case can be an outlier, distinguishable, or overruled. Keep going until the picture is stable.
• Trusting the headnote as the law. Headnotes are editorial summaries, not the court's words. Use them to navigate, then read the holding.
• Ignoring jurisdiction. A factually perfect case from the wrong court may not bind. Always confirm which court's law controls.
• Quoting dicta as a holding. A memorable line may be an aside the court did not need to decide the case. Confirm it is actually the holding before you rely on it.
• Skipping the good-law check. This is the mistake that undoes everything. A case that was solid last year may have been overruled since.
• Trusting an AI summary without opening the case. Summaries can be wrong or, worse, invented. Read the opinion and verify the citation.

Most research failures come down to one of these, and each one is avoidable with a little discipline.

How long should case-law research take?

There is no fixed answer, because a narrow question in a well-settled area can resolve in minutes while a novel question with sparse authority can take much longer. The point of a method is not to be fast for its own sake, but to be thorough efficiently: spend your time reading and evaluating the cases that matter rather than fighting the search itself. Plain-English tools help by shortening the search step, so more of your time goes to the reading, ranking, and verifying that actually determine whether your research is sound.

Staying organized

Research you cannot reconstruct is research you will redo. Keep a simple record: the issue, the searches you ran, the cases you kept and why, the holding of each, and the date you verified good law. Saved research and shared folders help teams avoid duplicating work, and they make it easy to update a memo when the law shifts.

From research to argument

Research does not end when you have a ranked list of good cases. The next move is turning that authority into an argument, which means showing how the holdings apply to your facts. For each case you rely on, be ready to explain why it is binding or persuasive, what its holding is, how its facts resemble yours, and why any contrary case can be distinguished. Good research makes this step easy, because you have already read the holdings and sorted the cases by authority. Weak research makes it painful, because you are scrambling to understand cases you only skimmed. The reading and ranking you do up front is what lets you write with confidence later, and it is why the method rewards patience in the early steps.

The honest boundary

A clean method produces good candidates, but it does not turn research into advice. AI-assisted and plain-English tools, including Cases, are for informational research only, not legal advice. Always read the opinions yourself and verify the citation at its source before relying on it. The method makes you faster and more thorough; your judgment and verification make the result trustworthy.

Key takeaways

• Frame a precise issue and the controlling jurisdiction before you search.
• Triage with headnotes, then read the holding, not the dicta.
• Expand backward and forward from your best case to find foundational and recent authority.
• Rank cases by authority and relevance, favoring binding decisions from your jurisdiction.
• Always verify each case is still good law and exists at its source.

Want to run this method faster? Try plain-English case law search with citations you can check, and see transparent pricing.