tldr: Good QA interview questions test judgment, not trivia. Anyone can memorize the difference between verification and validation; the signal is whether a candidate can decide what to test, what to automate, and when to stop. This is a question bank across manual, automation, and API testing.
How to read the answers
The sample answers below are what a strong candidate sounds like, not a script to grade against word for word. What you are listening for is reasoning: do they weigh risk, cost, and coverage, or do they recite definitions.
A weak candidate recalls facts. A strong one asks clarifying questions and explains tradeoffs.
Manual testing questions
How would you test a login form? Look for breadth: valid and invalid credentials, empty fields, SQL and script injection, lockout after failed attempts, password reset, session handling. A strong answer covers security and edge cases, not just the happy path.
When do you stop testing? There is no "when it's bug-free." Listen for risk-based reasoning: critical paths covered, exit criteria met, remaining risk understood and accepted by the team.
What is the difference between severity and priority? Severity is technical impact, priority is business urgency. A typo on the homepage is low severity, high priority. They should give an example like that unprompted.
Automation questions
What should you automate, and what should you not? Strong answer: automate stable, repetitive, high-value regression. Do not automate one-off checks, constantly changing UI, or exploratory work. The judgment about what not to automate is the real signal.
How do you handle flaky tests? Listen for root-cause thinking: fix the wait or the test data, do not just add a retry. Anyone who answers "increase the timeout" and stops has not maintained a real suite.
How do you keep a test suite maintainable? Good answers mention stable locators (roles and test ids over brittle CSS), shared helpers, and treating test code like product code. See Playwright interview questions for framework-specific depth.
API testing questions
What do you check in an API test? Status codes, response schema, data correctness, error handling, authentication, and edge inputs. A strong answer also mentions contract stability between versions.
How do you test for security at the API layer? Authentication and authorization (can one user reach another's data), input validation, rate limiting, and sensitive-data exposure in responses.
FAQs
What is the best single QA interview question?
"How would you test this?" on a concrete feature. It reveals how a candidate thinks about coverage, risk, and edge cases better than any definition.
Should QA candidates know how to code?
For automation roles, yes, at least enough to write and maintain test code. For manual roles, judgment about what to test matters more than coding.
How do I evaluate automation experience?
Ask what they chose not to automate and how they handled flaky tests. Both reveal real suite experience that definitions cannot.
Where can I find framework-specific questions?
See our Playwright interview questions page for a deeper, tool-specific bank.