#Environment variables

Rstest sets a small set of environment variables you can read from tests, setup files, and source code.

#NODE_ENV

Set to 'test' when the test runner starts (if not already defined). Existing values are preserved, so you can override it from your shell or package.json script.

Available as both process.env.NODE_ENV and import.meta.env.NODE_ENV across all test environments, including browser mode. Prefer import.meta.env.NODE_ENV in ESM source code where process may not be available.

if (import.meta.env.NODE_ENV === 'test') {
  // running under Rstest
}

#RSTEST

process.env.RSTEST is set to 'true' whenever the test runner is active. Use it to gate test-only branches in source code.

if (process.env.RSTEST) {
  // running under Rstest
}

For production builds, define process.env.RSTEST as false so the bundler can eliminate the dead branch — see Detect Rstest environment.

For in-source tests, prefer import.meta.rstest — it exposes the Rstest test API directly and is undefined in production builds.

#RSTEST_WORKER_ID

A stringified integer that uniquely identifies the worker process running the current test file. The first worker is '1'. Equivalent to Jest's JEST_WORKER_ID.

Use it to isolate shared external resources across parallel workers — typically database schemas, ports, temp directories, or any resource that would collide if two workers used the same name.

import { afterAll, beforeAll, test } from '@rstest/core';

const dbName = `myapp_test_${process.env.RSTEST_WORKER_ID}`;

beforeAll(async () => {
  await createDatabase(dbName);
});

afterAll(async () => {
  await dropDatabase(dbName);
});

test('inserts a row', async () => {
  // ...
});

Concurrent workers are always given different IDs, so the value is safe to use as part of a resource name during a worker's lifetime. The maximum number of concurrent workers is bounded by pool.maxWorkers; IDs are assigned monotonically as workers spawn and are not recycled within a single Rstest run.