$202,730top of the range in Washington · middle $104,300 / yr
High AI exposure
QA Testers in the United States earn a median of $104,300 a year. Pay starts near $61,440. Pay reaches $202,730 at the top of the range in Washington, the best-paying state for this work among those with at least 500 people in the job.
Source: U.S. Bureau of Labor Statistics, Occupational Employment and Wage Statistics, May 2025 (Software Quality Assurance Analysts and Testers, SOC 15-1253). Last checked 9 September 2026.
Entry level
$61,440
Top of the range · Washington
$202,730
Education
Bachelor's in CS or IT
Wages — U.S. Bureau of Labor Statistics, Occupational Employment and Wage Statistics, May 2025 (Software Quality Assurance Analysts and Testers). Top of the range is the highest state-level figure among states with at least 500 people in the job. AI-impact rating is PayCrunch's editorial assessment. Updated September 2026.
🆕 New & Trending AI Tools for QA TesterReviewed September 2026
We track new AI-tool launches every week and refresh this list — here’s what’s gaining traction for QA Tester work right now.
Claude CodeNEWFree / usage-based
Terminal coding agent that reads your repo, runs tests, and ships multi-file changes.
How a QA Tester uses it: describe a feature and let it implement and test it across the codebase
OpenAI CodexNEWIncl. w/ ChatGPT plans
Agent that runs longer, deterministic multi-step coding jobs on its own.
How a QA Tester uses it: delegate a well-defined build or migration and review the finished result
WindsurfNEWFree / $15 mo
Agentic IDE that keeps context across a whole project.
How a QA Tester uses it: make large, coordinated changes without losing track of the codebase
AWS KiroNEWPreview / see site
Spec-driven coding agent that turns written specs into working code.
How a QA Tester uses it: write the spec first and let it build to that spec
NotebookLMNEWFree / $7.99 mo
Google tool that answers questions grounded only in the documents you give it — with citations.
How a QA Tester uses it: load your own manuals, policies, or PDFs and ask questions that stay accurate to the source
CursorFree / $20 mo
AI-native code editor that edits across an entire project.
How a QA Tester uses it: describe a change in plain English and let it rewrite and refactor whole files
GitHub Copilot (Agent Mode)$10–19 mo
AI pair-programmer built into VS Code and GitHub that now completes multi-step tasks.
How a QA Tester uses it: hand off a task and have it plan, edit multiple files, and open a pull request
ChatGPTFree / $20 mo
The most-used AI assistant — writing, analysis, research, and images from a plain-language chat.
How a QA Tester uses it: draft emails and documents, summarize long files, and get instant answers to on-the-job questions
ClaudeFree / $20 mo
AI assistant known for careful writing, long-document analysis, and coding.
How a QA Tester uses it: analyze big reports or spreadsheets and turn messy notes into clean, finished writing
A QA tester sits with a build and tries to use it the way a person would. You follow a path through a screen. You watch what happens. You write a note when the product does something it should not. Then you say, before a release, which checks you actually ran and what you still have not seen. The work is hands-on. It is patience with a keyboard, a clear bug note, and a release conversation that does not pretend the unchecked parts were checked. Teams that skip that conversation ship surprises. Teams that keep it ship fewer of them.
Hands on a build
You start with what the team said the feature should do. A short description. A design. A sentence from the person who asked for it. Then you use the build. You try the obvious path, because customers will. You try the messy path, because customers will do that too: going back, leaving a field empty, using an old account, opening the same screen twice. You are not trying to be clever for its own sake. You are trying to see the product. A tester who only walks the happy path will bless a feature that falls apart the first afternoon it is real.
You keep a list of checks so the next build is not a fog. The list is yours, written in the language of the product: log in, save, edit, cancel, come back tomorrow. You mark what you ran on this build and what you did not. Honesty about the gap is part of the test. A green feeling is worthless if you cannot say which checks produced it. Developers and product managers can plan around a gap. They cannot plan around a shrug that said "looks fine."
Devices and accounts matter. The bug that appears only on one browser, or only for a person with an old profile, is still a bug. You note the setup. When time is short, you say which setups you covered and which you skipped. That record is more useful than a longer list of checks you only implied.
The bug note
When something breaks, you write it down so someone else can see it too. What you did. What you expected from the description of the feature. What the product did instead. The account or the screen, if that matters. A note a developer can follow is a gift. A note that says "it's broken" is a delay. You are not blaming the person who built it. You are pointing at a behavior. Testers who write with contempt get ignored even when they are right. Testers who write with precision get the fix, and they get asked to look at the next build.
You also say how bad it is in ordinary language. Can a person finish the task. Is the problem on the main path or on a corner. Does it lose information. You do not need a private scoring ritual to be useful. You need the reader to know whether this should hold a release. If you are unsure, say you are unsure and show what you saw. False certainty causes two kinds of damage: a release held for a trivial corner, or a release sent with a broken main path because the note sounded casual.
After a fix, you look again. The original path, and the nearby path the fix might have disturbed. You update the note. Closed means you saw the new behavior, not that someone told you it was closed. This is still hands-on work. A tester who only files notes and never returns to the build becomes a mailbox. The value is the second look, written just as clearly as the first.
A check you ran, a note someone can follow.
Say what you did and what you saw. A release conversation depends on that record, not on a feeling about the build.
Sitting in the release
Release day, or the meeting before it, is where your notes earn their keep. You tell the room what you checked, what failed and is still open, and what you did not have time to touch. You recommend from that record. Sometimes the recommendation is to wait. Sometimes it is to ship with a known problem, named in public, because the problem is small and the delay is costly. The recommendation is yours to make clearly. The decision may sit with a product owner. You should still be unwilling to smile at a gap you did not mention.
After the release you watch the first real use, within the access your job gives you. A spike in the same bug you wrote down is information. A new complaint in a corner you skipped is also information, and it belongs in the next round of checks. Testers who vanish at the moment of release miss the lesson. Testers who stay for the first feedback get better at choosing what to check when time is short. That judgment is the senior version of the job.
Keep this picture hands-on. The career described here is a person at the product, writing notes, speaking in the release meeting. It is not a tour of automated pipelines, build scripts, or a factory of jobs that run unattended. Some colleagues live in that work. This seat lives in the checks you can demonstrate. If an interview drifts into a lecture about tooling you have not used, bring it back to a bug you found with your own hands and a release you told the truth about.
Learning by checking
People become testers from support, from a junior software role, from a boot camp, or from a degree in computer science they discovered they would rather use this way. The preparation that counts is practice: bugs you wrote well, and a product you can talk through. There is no single license that makes you a QA tester. Some employers like a certificate from a testing board such as the ISTQB. That certificate proves you completed their program. It does not prove you can find a bug in their product. This note will not describe an exam. Bring the certificate if you have it, and bring the notes either way.
Learn the product you are hired to test more deeply than a casual user, and less arrogantly than the person who built it. Read the description. Use the feature until the odd corners feel familiar. Ask a developer what they are worried about, and then check that, plus the thing they are not worried about. Curiosity is the skill. Contempt for developers is a hobby that makes you worse at the job. You need them to read your notes. They need you to look at what they cannot see while they are building it.
Keep a portfolio with secrets removed. Three bug notes, rewritten so they reveal no customer data, plus a short account of a release call you were part of. What you had checked. What you flagged. What shipped anyway, and whether you had named it. That story tells a hiring manager you understand the seat. A list of tools tells them you have installed software. Tools change. A clear note survives the change.
Companies that still want a person at the keyboard
Product companies, contract studios, banks, hospitals, and public agencies all hire testers. The domain changes the checks. A payment screen and a clinic portal should not be tested with the same casualness, because the harm of a mistake is different. You do not need to be a banker or a clinician. You do need to learn which path is the main path in that domain, and which mistake would be serious. Say that in the interview. "I click around until something breaks" is a weaker story than "I learned what this product must get right, and I checked that first."
Ask who writes the description of the feature, who reads your notes, and who decides the release. Ask whether you are the only tester. A solo tester on a large product is a different job from a tester on a team with a lead who reviews your notes. Ask what "done" means before a release. If the answer is that testing time is whatever remains after development slips, believe them, and decide if you want that life. The wage should reflect a planned check, not a leftover afternoon.
Some postings mix this seat with automation, performance labs, or security work. Read the verbs. If they want hands-on checks, bug notes, and a voice in the release, you are in the right conversation. If they want a pipeline and a dashboard as the whole job, that is a different craft, and you should not pretend a week of manual checks was that craft. Honest scope is how you get hired into work you can do. It is also how you avoid being blamed for a factory you were never in.
Owning a release without leaving the checks
You begin by executing checks someone else listed, and by writing notes a senior edits. You learn what a useful note looks like in this company. The next step is choosing the checks: you decide what matters on a short build, you say what you skipped, and your recommendation in the release meeting is one the team uses. Later you may lead other testers. You review their notes, you protect time for the main paths, and you speak for the group when a date is unrealistic. Pay should move when the judgment moves. Running a release conversation on a junior wage is a reason to renegotiate.
Some testers later learn to script checks they already understand. That can be growth. It should not replace the hands-on skill, and it should not be required before you are trusted at the work described here. A tester who can find a serious bug and explain it still matters. Price the job you are doing this year.
Moving companies means learning a new product's main path. Your notes are portable. Your sense of the old product is not a wage by itself. The figures below are occupation statistics. They are a check on an offer. They are not a contractor's day rate, and they are not a bonus for a launch. Compare the base, then describe the checks you will own. That pairing is the negotiation.
Analysts and testers, with the states in order
The wages are Occupational Employment and Wage Statistics for May 2025, for Software Quality Assurance Analysts and Testers. The series name includes analysts as well as testers. The work in this note is the hands-on tester: checks, a bug note, a release. Hands-on testing pay on this series opens at $61,440. The national median is $104,300, a climb of $42,860. Washington is where the published top reaches $202,730. From the national median up to that high end is $98,430. Washington's median is a different statistic: $128,480. The highest median is California, at $128,740, which sits $24,440 above the national median. Washington's high end, Washington's median, and California's median are three different claims. Do not fold them into one "West Coast number."
State medians, in this order. Virginia's median is $121,590. Massachusetts's median is $122,210. Colorado's median is $123,690. Washington's median is $128,480. California's median is $128,740. Oklahoma's median is $77,700, the low end of this comparison. The gap between Oklahoma's median and California's median is $51,040. Virginia, Massachusetts, Colorado, and Washington sit in a band well above the national median, with California at the top of these medians by a small step over Washington. None of those medians is the $202,730 high end. That figure is the top of the published range in Washington, not Washington's median, and not California's median.
Keep the gaps labeled. $42,860 is entry to the national median. $24,440 is the national median to California's median. $98,430 is the national median to Washington's high end. $51,040 is Oklahoma's median to California's median. The distance between Washington's median of $128,480 and California's median of $128,740 is small beside the distance between Washington's median and Washington's high end of $202,730. If a recruiter quotes the larger Washington figure, ask whether they mean the top of the published range. If they mean typical pay in Washington, the median here is $128,480. If they mean the highest median on this list, that is California at $128,740.
Washington's high end beside California's median
An offer near $61,440 matches entry: you are running checks someone else chose, and a senior still rewrites your notes. If you already choose what to check, write notes developers use, and speak in the release meeting, and the offer is still at entry, name the $42,860 between entry and the national median of $104,300. An offer near $104,300 is the national middle for this series. It is a fair check on a working tester. In several of the states below, the local median sits higher, so use the place before you decide the middle is the whole story.
Match the median to the state, in order. Virginia at $121,590. Massachusetts at $122,210. Colorado at $123,690. Washington's median at $128,480. California's median at $128,740. In California you may name the $24,440 between the national median and that highest median. In Washington, keep the median and the high end apart: $128,480 is typical pay on this chart, and $202,730 is the high end of the published range, with $98,430 between the national median and that high end. Use the high end only for scope at the outer published reach, and say so. Oklahoma's $77,700 median, and the $51,040 spread up to California, show how far typical pay moves. They do not turn every testing seat into Washington's high end.
Benefits, a bonus, and whether you are the only person checking the product belong beside the base. Then match the offer to $61,440, to $104,300, to the state median in the order above, or to Washington's high end only when you mean the high end. Say $128,480 for Washington's median. Say $128,740 for California's median. Say $202,730 for the top of the published range in Washington. The series name belongs to the pay comparison. The job is still the build in front of you: a check, a bug note, and a release you described honestly.
The top of QA Tester pay — and how to get there with AI
$202,730what QA Tester pay reaches in Washington
Highest state-level top-of-range annual wage for Software Quality Assurance Analysts and Testers, among states with at least 500 people in the job. U.S. Bureau of Labor Statistics, Occupational Employment and Wage Statistics, May 2025.
And the role it leads to — Software Developers — reaches $272,670 in California.
$61,440entry$104,300middle$202,730top end
A QA tester in the middle of this range runs the test plan somebody else wrote; the one at the top carries a second credential — cloud, security, or the regulated domain being tested — and is trusted to say whether a build ships.
Most of the visible work here is repeatable: updating automated test scripts so they stay current, testing system modifications ahead of implementation, monitoring bug resolution until it closes. Assistants draft those scripts now and summarise a log file faster than a person reads one. What did not get cheaper is judgement about risk — sitting in a product design review and naming the schedule problem before it is built, or reviewing software documentation for compliance gaps that a test suite will never surface. A credential in the platform or the standard your product is held to is how an employer decides that judgement is real.
Your playbook, by where you are now
Just startingOwn the debugging trail
Perform your own initial debugging before filing anything: open the configuration files, read the logs, and name the breakdown source instead of handing over a screenshot.
Rewrite one flaky automated test script a week so the suite people ignore becomes the suite people trust.
Visit the beta testing sites your product uses and record how real performance differs from the staging build.
Keep a running defect notebook in Evernote or Airtable — what broke, what class of change caused it, how long resolution took.
Ask Claude or Gemini to summarise a long log into candidate failure points, then verify each one by hand before it reaches a developer.
What proves it: A defect log showing you identified breakdown sources, not just symptoms.
Realistic span: the first eighteen months to two years
A few years inAdd the credential your product needs
Pick the credential from the stack you already test against — the cloud provider behind your Amazon Elastic Compute Cloud EC2 and Amazon DynamoDB services, or the security standard your compliance team quotes.
Study it against live work: every practice question should map to something in your own product's Amazon Redshift reporting or ADO.NET data layer.
Once certified, review software documentation for compliance and completeness as a named duty, not a favour.
Plan test schedules and strategies against real delivery dates, and publish the risk you accepted when a date forced a cut.
Give developers written feedback on usability and functionality with the standard cited, so the recommendation carries weight past your desk.
What proves it: The certification plus a compliance review of your own product that changed a release decision.
Realistic span: years three through six
ExperiencedSit in design review, not just the queue
Join product design reviews early and raise functional requirements, schedule pressure and likely failure modes while they are still cheap to change.
Set the test strategy for a whole product line and defend it to delivery managers who want the schedule shortened.
Build the packaging and install checks — an Acresso InstallAnywhere run counts — so the first customer experience is tested like everything else.
Train the newer testers on the credential you hold and on how to read a configuration file properly.
Move into development if you want the wider scope; Washington employers pay QA testers the most, and the developer track sits directly above this one.
What proves it: A test strategy for a product line that other testers execute and delivery managers plan around.
Realistic span: seven years and onward
The next 90 days
Take your last twenty defects and mark which ones you found by executing a written step and which ones you found by reading a log, a configuration file or a specification and noticing something wrong. The second group is the part of QA testing that an automated suite cannot reproduce, and it is the part that argues for the credential. Look at what those defects touched — a cloud service, an access-control rule, a regulated data field — and register for the certification covering that area within the next ninety days. While you study, review one piece of your product's documentation against that standard and write down every gap you find. Bring both the gap list and the exam result to your next review conversation.
Wage figures: BLS OEWS, May 2025. The playbook is PayCrunch editorial guidance, not a guarantee of pay or placement.
Every figure is the national median from the U.S. Bureau of Labor Statistics (OEWS) shown on that role’s own page.
Never used AI before? Start here (2 minutes).
Start by making AI write your test cases - it's the fastest coverage win. Open ChatGPT or Claude, paste a de-identified user story or acceptance criteria, and ask for a full set of test cases including edge and negative paths. You'll immediately catch scenarios you'd have missed, and you'll build the habit of thinking in coverage, not clicks.
Then start escaping manual regression: pick a codeless AI automation tool like testRigor, Mabl, or Katalon and automate one repetitive test flow this week. Learning automation is the single biggest pay lever in QA. Keep real customer data out of every AI tool and every test environment. AI is the junior tester who drafts and executes; you decide what 'done' and 'safe to ship' actually mean.
The one rule, forever: Never paste production data, customer PII, credentials, or secrets into a consumer AI tool - use synthetic or masked test data, and keep testing to authorized, non-production environments. AI-generated tests create false confidence: a green suite isn't proof of coverage, so review what the AI actually asserts. You own the quality sign-off; the AI doesn't catch what it wasn't told to look for.
The plays — exact steps, exact prompts
Do these in order. Each one is copy-paste ready. You do not need to know anything about AI going in.
1
Generate complete test coverage from requirements with AI
Why this pays: Coverage gaps are where bugs escape and QA reputations die. AI that turns a requirement into an exhaustive test set - including the edge cases you'd forget - makes you visibly more thorough, the reliability that earns senior QA pay.
ChatGPTClaudeTestRail
1
Paste a de-identified user story or acceptance criteria into ChatGPT or Claude and ask for positive, negative, boundary, and edge-case tests, then manage them in TestRail.
2
Prompt for structured, thorough coverage.
Copy-paste this prompt
You are a senior QA engineer. For this feature, write a complete set of test cases: happy path, negative paths, boundary values, and edge cases (empty input, max length, special characters, concurrency, permission variations). Output a table with Test ID, Preconditions, Steps, Expected Result, and Priority. Feature: [paste acceptance criteria]. Also call out any ambiguous requirement that needs clarification.
Use only non-confidential requirements. Review the AI's cases - it invents plausible-looking tests that may not match real system behavior.
3
Add the domain-specific and integration cases the AI can't know about from the requirement alone.
What you'll haveBroader, sharper coverage on every feature - the thoroughness that gets you trusted with critical releases.
2
Escape manual regression with codeless AI automation
Why this pays: Automation is the clearest dividing line between a manual tester and a six-figure QA engineer. Codeless AI tools let you automate regression without deep coding - the skill jump that directly moves your pay toward $202,730.
testRigorMablKatalon
1
Pick a repetitive regression flow and automate it in testRigor (plain-English tests), Mabl, or Katalon - no framework code required to start.
2
Use AI to draft the test steps in the tool's syntax.
Copy-paste this prompt
Write a testRigor test in plain-English commands for this user flow: [log in, add an item to cart, apply a promo code, check out, verify order confirmation]. Include an assertion at each step and a negative test for an invalid promo code. Use [test-account placeholders], never real credentials.
Never automate against production with real accounts. Verify each generated step actually validates the intended behavior.
3
Grow your automated suite one flow at a time until regression runs hands-off in CI.
What you'll haveA regression suite that runs itself and a genuine automation skill on your resume - the pay-grade jump out of manual QA.
3
Make suites resilient with visual and self-healing AI
Why this pays: Flaky, high-maintenance automation is where QA credibility leaks away. AI visual testing and self-healing locators cut maintenance and catch UI regressions humans miss - the reliability that makes you the automation lead.
ApplitoolsTestimPercy
1
Add Applitools visual AI to catch layout, rendering, and cross-browser regressions that assertion-based tests skip.
2
Use Testim's self-healing locators so tests survive minor DOM changes instead of breaking on every deploy.
3
Audit AI-approved visual diffs so real regressions aren't rubber-stamped as expected.
Copy-paste this prompt
Review this list of visual-diff results from an automated UI test run: [paste summary of changed regions]. For each, reason about whether it's likely an intentional design change or a genuine regression, and which ones I should manually verify first. Explain your reasoning.
AI baselines can auto-accept a real bug as the new normal - spot-check the diffs, especially on critical screens.
What you'll haveLow-maintenance, high-signal suites that catch what others miss - the reliability that earns the automation-lead role.
4
Supercharge exploratory testing and bug reports with AI
Why this pays: Great exploratory testing and crisp, reproducible bug reports are what developers and managers remember. AI helps you generate test charters, analyze logs, and write flawless repro steps - the craft that builds a standout QA reputation.
ChatGPTClaudeJira
1
Before an exploratory session, have AI generate a test charter and heuristics tailored to the feature and its risks.
2
Turn a rough observation into a perfect bug report.
Copy-paste this prompt
Turn my rough notes into a clear, reproducible bug report for Jira: Title, Environment, Preconditions, numbered Steps to Reproduce, Expected vs Actual, Severity and Priority with justification, and suggested logs or attachments to include. Notes: [paste rough notes]. No customer data.
Strip any real user data from logs and screenshots before sharing. A reproducible report is your credibility - verify the steps actually reproduce it.
3
Paste de-identified error logs or stack traces into AI to hypothesize root cause and narrow the repro.
What you'll haveSharper exploratory sessions and bug reports developers act on immediately - the craft reputation that gets you promoted.
5
Generate test data and automate API testing
Why this pays: Realistic test data and solid API coverage are perennial bottlenecks. AI that generates data and API tests removes them - expanding what you can test and adding backend skills that command higher QA pay.
Postman (Postbot)ChatGPTFaker
1
Use Postman's Postbot AI to generate API tests, assertions, and documentation from an endpoint, then chain them into a collection run.
2
Generate rich synthetic test data.
Copy-paste this prompt
Generate 30 rows of realistic but entirely synthetic test data for a [user registration] form as CSV: fields are [name, email, phone, date_of_birth, country, postal_code]. Include valid rows plus deliberate edge cases (max-length names, unusual but valid emails, boundary dates, international formats, and clearly invalid entries for negative testing). Never use real people.
Use only synthetic data - never real customer records. Confirm the edge cases match your system's actual validation rules.
3
Build API test coverage that runs in CI so backend regressions are caught before the UI.
What you'll haveFaster test-data setup and real API coverage - broader testing skills that push you up the QA pay band.
6
Move from tester to quality strategist
Why this pays: The top of QA pay goes to those who own quality strategy or a specialty (performance, security, accessibility), not those who execute scripts. AI accelerates the learning and the leadership artifacts that get you there.
NotebookLMaxe DevToolsChatGPT
1
Pick a direction - automation architecture, performance, security, or accessibility - and use NotebookLM to master its core body of knowledge.
2
Draft a quality strategy that gets you noticed.
Copy-paste this prompt
Act as a QA lead. Draft a one-page quality strategy proposal for a team shipping [a web app] every two weeks: the test pyramid and what to automate at each level, entry and exit criteria for releases, the key quality metrics to track (escape rate, coverage, flake rate), and a 90-day plan to reduce escaped defects. Generic best-practice, no confidential details.
Adapt the strategy to your team's real stack and constraints. AI gives the framework; the judgment about your context is yours.
3
For accessibility, pair axe DevTools with AI to find, explain, and prioritize WCAG issues - a specialized, well-paid niche.
What you'll haveA path from executing tests to owning quality - the strategic and specialist roles that reach the top of the pay band.
Your 12-month sequence to the top of the range
How the plays above stack into a path from median pay toward the $202,730 tier.
Month 1
Use AI (ChatGPT or Claude) to generate thorough test cases and write flawless, reproducible bug reports.
Months 2-3
Automate your first regression flows with codeless AI tools (testRigor, Mabl, Katalon).
Months 3-6
Harden suites with visual and self-healing AI (Applitools, Testim) and add API testing (Postbot).
Months 6-9
Pick a specialty - automation, performance, security, or accessibility - and go deep with AI's help.
Months 9-12
Own a quality-strategy initiative for your team - the leadership route to the top pay band.
Next steps for a QA Tester
Some links below are affiliate or partner links. PayCrunch may earn a commission if you enroll or subscribe through them, at no extra cost to you. Wage figures on this page still come from the Bureau of Labor Statistics, not from these programs.
QA Tester work is specific enough that a stamped 'check out these courses' block would be noise. BLS files this work as Software Quality Assurance Analysts and Testers (SOC 15-1253). O*NET Job Zone 4 is typical: a bachelor's degree, so the honest next credential is a professional certificate or bachelor's-level coursework — not a random catalog dump.
The occupation's listed knowledge areas include Engineering and Technology and Design; the links search those subjects, not a generic 'career courses' list.
QA Testers in this dataset list AJAX among the tools in use, so a program that names that stack is a better fit than a survey course.
Coursera search for engineering and technology — a professional certificate or bachelor's-level coursework that lines up with computing, not a generic professional-development aisle.
FlexJobs screens remote, hybrid, freelance, and flexible listings so you are not wading through unverified ads. This is a job-board search for QA Tester work, not a claim that they list a counted SOC 15-1253 inventory.
Write a QA Tester resume, or one aimed at Software Developers, instead of a blank template. Resume Now is a resume builder; we are not claiming a counted template set for this SOC.
A QA Tester resume that names the actual tasks on this page, or the step-up title Software Developers, beats a blank template when you apply.
What QA Testers earn by state
These are the Bureau of Labor Statistics’ own figures for Software Quality Assurance Analysts and Testers, state by state — not a cost-of-living adjustment applied to the national number. Only states employing at least 500 people in the occupation are shown, because a state median drawn from a handful of workers is noise rather than a signal.
California
$128,740
highest of them · +23% vs the national median
Oklahoma
$77,700
lowest of the 35 states and D.C. that qualify · -26% vs the national median
The same job pays $51,040 more a year at the median in California than in Oklahoma — 66% higher. That gap is what the Bureau measured, before any question of what it costs to live in either place. The top-of-range figure quoted at the head of this page, $202,730, is a different statistic in a different place: it is the 90th-percentile wage in Washington. The state that pays the typical worker most and the state where the best-paid go highest are not always the same one.
Source: U.S. Bureau of Labor Statistics, Occupational Employment and Wage Statistics, May 2025, SOC 15-1253. 35 states and D.C. clear the 500-employee reporting floor for this occupation; those below it are left out rather than shown with a wide error band.
Free data. Use any of it.
PayCrunch publishes verified, BLS-sourced salary + AI-playbook data on 1,000+ professions — free, no signup.
The repetitive, manual, script-following part of QA is genuinely being automated - AI can generate tests, self-heal them, and run regression, so pure manual click-testing is a shrinking, exposed role. But someone still has to decide what to test, judge whether a green suite really means quality, own risk-based release decisions, and do the exploratory and specialized testing AI can't. QA testers who move toward automation and strategy are safer and better paid; those who stay manual-only are among the most exposed roles on this list.
Can I trust AI-generated tests?
Only after you review them. AI writes plausible-looking test cases and automation fast, but it invents assertions that may not match real system behavior, and a passing suite it generated can give false confidence while missing the actual bug. Treat AI tests as a strong first draft: verify each one validates what it claims, and add the domain and integration cases the AI couldn't know about.
How does AI move a QA tester's pay toward the top?
By moving you off manual execution and onto higher-value work. AI removes repetitive test writing and running, which frees you to learn automation (the biggest single pay jump in QA), specialize in performance, security, or accessibility, and own quality strategy. Automation and strategy skills are what separate a low-paid manual tester from a $202,730 QA engineer.
Is it safe to use ChatGPT for testing work?
For non-confidential requirements, test-case generation, synthetic data, and log analysis - yes. But never paste production data, customer PII, credentials, or secrets into a consumer tool, and never test against production with real accounts. Use masked or synthetic data and authorized environments; the quality sign-off and the data handling are your responsibility.
Which AI skill should I build first?
Two, in order: AI test-case generation (immediate coverage wins with zero setup), then codeless AI automation with testRigor, Mabl, or Katalon (the skill that actually raises your pay grade). Start generating test cases this week and automate your first regression flow this month - that combination moves you from manual tester toward QA engineer fastest.
Methodology & sources
Salary (median, 10th, top of the range) — U.S. Bureau of Labor Statistics, OEWS.
By state — the Bureau of Labor Statistics’ own state medians, limited to states employing at least 500 people in the occupation. No cost-of-living arithmetic is applied to a wage anywhere on this page.
The plays — PayCrunch's own step-by-step guidance using publicly available AI tools. Tool names/URLs are real and current as of August 2026; prompts written to work as-is. Verify any professional output before relying on it.