roadmap
- 01
- 02
- 03
- 04
- 05
- 06
- 07
- 08
Prepare for Selenium interview questions on driving real browsers through WebDriver. Covers locators and waits, page objects, windows and frames, evidence capture, and scaling suites on Grid.
265 of 265 questions in your plan: 0 known, 0 in progress, 0 due, 0 skipped, 265 new. 0 cut, outside your plan.
Selenium comes up in QA and SDET interviews, but the two weigh this tree differently. In a QA engineer interview, frequently for a manual tester's first automation role, the questions stay at the browser: why a step that is green on your machine goes red on the CI agent, which locator survives a redesign, why a click lands on an overlay and throws `ElementClickInterceptedException`, what a `StaleElementReferenceException` says about the page between two lines of the test. An SDET loop treats Selenium as one library inside a framework the candidate is expected to build and defend, usually with a coding round attached, so the questions climb a layer: how a parallel suite keeps one driver per thread, where waiting belongs once page objects hide the elements, how a remote session is matched to a node by its capabilities, and what a failed nightly run leaves behind to debug. Sooner or later both ask why a team would still pick Selenium over a newer browser tool, and the answer they want weighs what the WebDriver protocol costs against what it buys. If the interview is for the job rather than the tool, the role roadmaps carry what this one leaves out: [Java SDET](/roadmaps/java-sdet) and [QA Engineer](/roadmaps/qa-engineer) set this topic beside the test design, language and tooling questions those loops spend most of their time on.
Each question here is meant to be answered before its answer is read, one at a time. The explanation that ties the subtopics to each other is [the Selenium primer](/topics/dev-selenium) on the topic's hub page; skim it before you start drilling, and reopen it when one answer seems to undo another. Juniors are judged first on [synchronization](/topics/dev-selenium-waits): why `Thread.sleep` is the wrong fix for a flaky step, what an explicit wait polls for, and why the Selenium documentation warns against mixing implicit and explicit waits. At middle level the deciding subject is the [abstraction layers](/topics/dev-selenium-page-objects): what a page object should return, whether assertions belong in it, why a `@CacheLookup` field goes stale after the page re-renders. Senior candidates should reach [WebDriver's architecture](/topics/dev-selenium-driver) early, because each classic WebDriver command is an HTTP request to a driver process, and that one fact explains the latency, the remote runs and most of the comparison with newer tools. Where Selenium is one tool among several on the posting, the page can stay almost as it opens: study order on by frequency, importance ending at should know. For an SDET loop with a framework round, switch to by topics, which reaches the protocol well before the design questions that rest on it.
Selenium comes up in QA and SDET interviews, but the two weigh this tree differently. In a QA engineer interview, frequently for a manual tester's first automation role, the questions stay at the browser: why a step that is green on your machine goes red on the CI agent, which locator survives a redesign, why a click lands on an overlay and throws ElementClickInterceptedException, what a StaleElementReferenceException says about the page between two lines of the test. An SDET loop treats Selenium as one library inside a framework the candidate is expected to build and defend, usually with a coding round attached, so the questions climb a layer: how a parallel suite keeps one driver per thread, where waiting belongs once page objects hide the elements, how a remote session is matched to a node by its capabilities, and what a failed nightly run leaves behind to debug. Sooner or later both ask why a team would still pick Selenium over a newer browser tool, and the answer they want weighs what the WebDriver protocol costs against what it buys.
If the interview is for the job rather than the tool, the role roadmaps carry what this one leaves out: Java SDET and QA Engineer set this topic beside the test design, language and tooling questions those loops spend most of their time on.
Each question here is meant to be answered before its answer is read, one at a time. The explanation that ties the subtopics to each other is the Selenium primer on the topic's hub page; skim it before you start drilling, and reopen it when one answer seems to undo another.
Juniors are judged first on synchronization: why Thread.sleep is the wrong fix for a flaky step, what an explicit wait polls for, and why the Selenium documentation warns against mixing implicit and explicit waits. At middle level the deciding subject is the abstraction layers: what a page object should return, whether assertions belong in it, why a @CacheLookup field goes stale after the page re-renders. Senior candidates should reach WebDriver's architecture early, because each classic WebDriver command is an HTTP request to a driver process, and that one fact explains the latency, the remote runs and most of the comparison with newer tools.
Where Selenium is one tool among several on the posting, the page can stay almost as it opens: study order on by frequency, importance ending at should know. For an SDET loop with a framework round, switch to by topics, which reaches the protocol well before the design questions that rest on it.