QA Foundations
Defect escape rate
Also known as: escaped defects
Defect escape rate is the percentage of observed defects first found after release, out of all defects found for the same release and observation window.
Escape rate (%) = 100 × E / (P + E). P is the number of distinct eligible defects first found before release, including reviews and testing. E is the number first found after release for the same release or cohort, within a declared post-release observation window. For example, 90 pre-release defects and 10 post-release defects give an observed escape rate of 10%.
Count confirmed product defects once, not duplicate reports, repeated test failures or infrastructure errors. Record severity and keep the release scope, defect classification and observation window consistent when comparing results. This convention measures discovery timing; track known-but-unfixed defects shipped at release separately.
NASA’s 1991 Software Engineering Laboratory report gives a nominal 0.5 errors per developed thousand source lines during maintenance/operations for its historical environment. That is defect density, not an escape percentage or a current industry benchmark.
A rising escape rate with a stable test pass rate is a reason to investigate, not proof that the suite is testing the wrong things. Discovery timing, reporting and release scope can change the ratio. Review the escaped defects, their severity and the missed scenarios alongside the counts.
CI pass rates and Qualflare’s pass-rate-based launch risk describe the tests that ran. They do not supply the pre- and post-release defect records needed for this calculation.
- With no observed defects (P + E = 0), the rate is undefined, not 0%.
- No observed escapes in the declared window does not prove that a release has no latent defects.
- Tracing an escaped defect to the change that introduced it helps prevent recurrence, but is not required to calculate this ratio.
Related terms
Sources
See it in your own test results
Qualflare detects flaky tests, clusters failures by root cause, and scores release risk from the test results you already produce in CI. Start free.
Start free with Qualflare← Back to the testing & observability glossary.
Last reviewed November 14, 2026