Test plans

A test plan gathers the test cases you intend to run into one defined set. Where a test case says how to verify one thing, a plan says what we’re going to verify in a given round of testing.

What a plan is for

Authoring test cases and deciding which to run are two different activities. A plan separates them: you build a library of reusable test cases over time, then assemble the relevant ones into a plan for a specific purpose — a feature, a release candidate, a regression sweep. The plan is the agreed scope, so everyone knows what’s in and what’s out.

Building a plan

You add the test cases that belong in the plan, organizing them into the order and grouping that make sense for how they’ll be run. Because test cases are reusable records, the same case can appear in more than one plan without being duplicated — a smoke test can live in both a release plan and a nightly plan, and stays a single source of truth.

Plans, cases, and runs

A test plan is a planning record, distinct from the test cases it references and from the runs that execute it. The plan defines the scope; a run is an actual execution of that scope, recorded with its own results. That separation lets you run the same plan many times — each run a fresh, independently recorded pass — which is exactly what you need across iterations and releases. Running a plan is covered next in Runs & execution.