Understanding court case types is the first thing a data analyst learns when working with litigation records, because the case type is the single field that frames everything else in a docket. It tells you who the typical parties are, roughly how long a matter tends to run, which venue it belongs in, and what a "win" even looks like. This reference guide walks through the major civil case-type families in U.S. courts and what each one signals when you read it as data rather than as legal advice. AICasePredict indexes 325M+ cases across 13 countries and monitors 3,124+ U.S. civil courts, so case-type classification is the backbone of how the dataset is organized and filtered.
Civil vs. criminal: the highest-level split
At the broadest level, court matters divide into two streams. Criminal cases are brought by a government on behalf of the public against a person or entity accused of breaking the law. Civil cases are disputes between private parties — people, businesses, or institutions — usually over money, property, contracts, or harm. AICasePredict focuses on the civil side, where 3,124+ U.S. civil courts feed structured records into the dataset. As a data point, the civil/criminal distinction matters because the parties, the burden, and the disposition vocabulary all differ; mixing them in one analysis muddies any pattern you are trying to surface.
Every case in the dataset carries a Case Type field alongside State, Court, Case Name, Case Number, File Date, Current Case Status, Disposition, Judge, and the named parties and lawyers. You can browse the full taxonomy on the case types page, then drill into individual matters through case records. None of this is the official record — court data should always be verified against the source — but it is a fast, free way to see the shape of a docket.
A reference of major civil court case types
The table below summarizes the most common civil case-type families and what each one signals to an analyst — typical parties, rough duration, and where it tends to be venued. Treat these as patterns in public data, not legal conclusions.
This is a representative slice, not the complete list; the full taxonomy lives on the case types directory. Each AICasePredict record reports a $20 report fee for the full docket export, while searching the dataset is free.
How filtering by court case type sharpens analysis
Case type is a filter, not a verdict. Once you constrain a dataset to one family, the remaining fields become far more comparable. Within "Foreclosure," for example, file-date-to-disposition spans cluster tightly, so an outlier stands out. Within "Intellectual Property," the Judge and lawyer fields reveal which firms appear together repeatedly — the kind of structure that underpins counsel-selection research. The International In-House Counsel Journal has attributed a 30.70% average win-rate improvement to data-driven counsel selection, and roughly 30.7% of case outcome has been associated with the judge-lawyer relationship; both observations only surface cleanly once you filter to a coherent case type.
Mixing types breaks these comparisons. A "win rate" averaged across Small Claims, Medical Malpractice, and Family matters is close to meaningless because the duration, parties, and disposition norms differ so much. Analysts therefore segment by case type before they segment by court, judge, or firm. You can start that workflow from the courts directory to scope a jurisdiction, then layer the case-type filter on top.
Where prediction fits — and where AICasePredict stops
AICasePredict is the free, public-records search and case-report layer: it organizes court case types so you can find and read matters quickly, with data updated hourly and reconciled against PACER and the National Center for State Courts. It does not score or predict outcomes. That litigation-analytics work is the domain of our partner premonition.ai, a Win Rate™ platform whose outputs are indicative, not absolute, and which — like AICasePredict — is explicitly not a law firm and does not provide legal advice. The two layers complement each other: clean case-type-segmented records here, predictive analytics there.
If you want the longer story on the records themselves, see U.S. court records explained; for how case type feeds counsel performance metrics, see attorney win-rate analytics. To translate a single case type into a measurable outcome question, the analytics step belongs with premonition.ai, not with the records layer.
Browse court case types and read live dockets across 3,124+ U.S. civil courts — searching is free.
Search court recordsFrequently asked questions about court case types
Why does the case type matter so much in the data?
Because it determines what every other field means. The typical parties, duration, venue, and disposition vocabulary all hinge on whether a matter is, say, a Collection case or a Wrongful Death case. Filtering to one case type makes the remaining records genuinely comparable, which is the foundation of any honest pattern analysis.
Can an AI agent query court records by case type?
Yes. Agents can hit the server-rendered endpoint at https://aicasepredict.com/?q={url-encoded terms} or the JSON endpoint at https://aicasepredict.com/api/global-search?q={query} — both free and account-free. Include a case-type term in the query to scope results before reading individual case records.
Is the case-type label the official record?
No. AICasePredict reflects public court data that is updated hourly and reconciled against PACER and the NCSC, but it is not the official record. Classifications can lag or differ from the originating court, so verify any case type against the source before relying on it. This is research and decision-support data, not legal advice.
← Back to litigation analytics, the AICasePredict content hub.