Testing Interview Questions

Frontend testing interview map — Testing Library, pyramid tradeoffs, mocks, a11y tests, and how to talk test strategy.

intermediate3 min readUpdated Aug 4, 2026

interview
testing-interview

Testing interviews rarely want 90% coverage brags. They want risk-based strategy: what you test, at which layer, and how you avoid brittle suites.

What interviewers score

Pyramid judgment — unit / integration / e2e balance
User-centric queries — Testing Library philosophy
Async & mocks — network without flaking
What not to test — implementation details
CI reality — flake control, a11y gates

Pyramid & tradeoffs

Layer	Best for	Cost	Learn
Unit	Pure logic, reducers, parsers	Low	Unit testing pure logic
Component / integration	User flows in JSDOM	Medium	Integration testing UI · Testing Library
E2E	Critical money paths	High	E2E tradeoffs · Playwright vs Cypress

Overview: Testing pyramid for frontend.

Out loud: “Many tests around pure pricing logic; RTL for form submit; one Playwright smoke for checkout.”

Testing Library core

Topic	Can you…	Learn
Query priority	role/label over test ids	Queries priority
user-event	Prefer over fireEvent	user-event vs fireEvent
Async	`findBy`, waitFor
test ids	Last resort	test ids

// Strong: role + name
await user.click(screen.getByRole("button", { name: /save/i }));
expect(await screen.findByText(/saved/i)).toBeInTheDocument();

Network mocking

Approach	When	Learn
Mock fetch	Small unit of client	Mocking fetch
MSW	App-level integration	MSW
Contract tests	Multi-team APIs	Contract testing

Flakes & timers

Topic	Can you…	Learn
Flake causes	races, shared state, time	Flaky tests
Fake timers	debounce/throttle tests	Fake timers
Snapshots	rare, intentional	Snapshot testing

Coverage & CI

Topic	Can you…	Learn
Coverage pitfalls	100% ≠ quality	Coverage pitfalls
A11y in CI	axe checks + limits	A11y testing CI · axe limits
Visual regression	design systems	Visual regression
Storybook	component states	Component testing Storybook

Classic questions

Question	Strong direction
Unit vs integration?	Isolation speed vs confidence in wiring
Why not test internals?	Refactors break tests; users don’t call hooks directly
How test hooks?	Via component or renderHook carefully
E2E count?	Few critical paths; rest lower pyramid
How handle auth in e2e?	Test user / storage state / stub IdP
Snapshot everything?	No — noise and rubber-stamp reviews

Strategy prompt (use structure)

“Design a test plan for an autocomplete.”

Unit: filter ranking function
RTL: type query, see options, keyboard select
MSW: slow/error network states
E2E: optional one path if search is revenue-critical
A11y: axe + manual keyboard

Practice UI: Autocomplete.

Footguns

Testing Redux connected guts instead of behavior
Snapshot of entire page HTML
Sleeping 5s instead of waiting for condition
Mocking so much nothing is real
Zero tests on pure money logic