Free tools Windows power users keep installed
One-click scans. No signup required.
Software testing techniques help you turn requirements, code paths, and risks into a small, defensible set of test cases. The practical choice depends on what you know about the system, which failures matter, and what you need to cover: specified behavior, internal code structure, or risks suggested by experience. No single technique finds every kind of defect, so combine methods when their coverage complements each other.
What are software testing techniques?
Testing techniques are systematic ways to analyze what to test and design test cases. They help avoid relying on a handful of arbitrary examples, while keeping the test set focused enough to execute and maintain. The International Software Testing Qualifications Board (ISTQB) Certified Tester Foundation Level (CTFL) Syllabus v4.0, dated 2023-04-21, groups core techniques into black-box, white-box, and experience-based approaches. It also covers collaboration-based approaches for making requirements testable.
A useful starting point is to name the test basis—the information from which you derive tests. That might be a requirement, a business rule, source code, a state model, a checklist, defect history, or a combination. Then decide what coverage item you need to exercise, such as input partitions, rule combinations, transitions, statements, or branches.
How do you choose a testing technique?
Start with the question your tests must answer, not with a favorite technique. Use the technique whose test basis and coverage item match the risk you are trying to reduce. The table is a selection aid, not a universal ranking of effectiveness or cost.
Crashes, No Sound, or Screen Glitches?
Random freezes, missing sound and display glitches usually trace back to one bad driver. Find and replace yours safely.Free scan · under a minutePC Slower Than It Used to Be?
A free scan shows the junk files, broken settings and background clutter dragging Windows down - then fixes them in one click.Free scan · Windows 10 & 11| Technique family | Test basis | Useful for | Coverage item | What you need |
|---|---|---|---|---|
| Black-box | Specified behavior, rules, interfaces, or acceptance criteria | Input classes, limits, rule combinations, and event sequences | Partitions, boundaries, rules, or transitions | A sufficiently clear behavioral basis and representative data |
| White-box | Internal control flow or code structure | Checking that implementation paths execute, including paths not apparent from requirements alone | Statements or branches | Access to code or a control-flow view and an agreed coverage measure |
| Experience-based | Tester knowledge, domain understanding, checklists, and defect history | Probing likely mistakes and discovering unexpected behavior | Depends on the charter or checklist; not necessarily a single formal coverage unit | Tester skill, system context, and a way to record observations |
| Collaboration-based | Shared examples, user stories, and acceptance criteria | Clarifying expected behavior before implementation | Agreed examples and acceptance conditions | Participation from people who understand the user need and business rules |
For each candidate technique, ask: What defect pattern am I targeting? What item will count as covered? Is the specification, code, data, or environment reliable enough? What expertise and maintenance will the tests require? Avoid comparing techniques using an unspecified phrase such as “test coverage”; name the unit being counted.
How do black-box testing techniques work?
Black-box test cases are derived from specified behavior rather than implementation details. If the code changes but the required behavior does not, these tests can remain useful. They are especially helpful when verifying requirements, business rules, public interfaces, and acceptance criteria. CTFL v4.0 material presented by ASTQB describes equivalence partitioning, boundary-value analysis, decision-table testing, and state-transition testing as black-box techniques.
Equivalence partitioning: sample behavior classes
Equivalence partitioning (EP) divides input values into groups expected to receive the same treatment. Select at least one representative from each relevant group. Include valid and invalid groups when they produce distinct behavior. Partitions should be non-empty and non-overlapping; real systems may need several partitions because different rules apply to different subsets.
For a checkout form that accepts a quantity from 1 through 5, possible partitions include valid quantities (1–5), a quantity below the minimum, a quantity above the maximum, malformed input such as text, and missing input. The exact partitions depend on the actual requirements: if zero means “remove item,” it is not equivalent to an arbitrary invalid quantity. EP reduces redundant sampling, but one representative per group does not prove every value in that group behaves correctly.
Boundary-value analysis: test the edges
Boundary-value analysis (BVA) focuses on edges of ordered partitions, where a limit may be shifted, omitted, or implemented with the wrong inclusive/exclusive operator. For the stated inclusive range 1–5, a three-value boundary set checks 0, 1, and 2 at the lower edge, and 4, 5, and 6 at the upper edge. This is the three-value convention: immediately below, at, and immediately above each boundary. A two-value method uses the boundary and its adjacent value; state which method your test plan uses.
Use values that are actually adjacent in the input domain. For integers, those are straightforward; for dates, decimals, or strings, establish the domain’s precision and ordering first. If the rule is exclusive—greater than 1 and less than 5—the expected behavior at 1 and 5 differs, so record the endpoint convention explicitly rather than reusing an inclusive-range expectation.
Decision tables: cover rule combinations
Use decision-table testing when combinations of conditions determine an outcome, especially for business rules. List meaningful conditions in rows, then create columns for condition combinations and the resulting actions. Derive tests to cover the applicable rules; combine columns only when their conditions genuinely lead to equivalent outcomes.
| Account active? | Payment valid? | Item in stock? | Expected outcome |
|---|---|---|---|
| Yes | Yes | Yes | Accept checkout |
| No | Yes | Yes | Reject: account inactive |
| Yes | No | Yes | Reject: payment invalid |
| Yes | Yes | No | Reject: item unavailable |
This table illustrates independent rejection rules; it is not a complete specification for every possible combination. If the application has precedence rules, such as returning a particular error when multiple conditions fail, add those combinations and expected messages explicitly. A decision table often reveals missing or conflicting requirements before they become test failures.
State-transition testing: exercise sequences
State-transition testing models the states a system can occupy, the events that move it between states, optional guard conditions, and resulting actions. Tests can cover valid transitions and, where relevant, invalid transitions or event sequences. A state diagram is useful for communicating the model; a transition table is useful for checking that each event/state combination has an expected result.
For a login workflow, model states such as active, locked, and reset pending. Events might include a successful login, a failed attempt, reaching the specified attempt limit, requesting a reset, and completing recovery. Test sequences, not just isolated screens: several failed attempts may lead to a lock, and a reset event may or may not restore access depending on the requirements. Include invalid sequences—for example, attempting a normal login while locked—when the system must define their outcome. Do not invent a lockout threshold; take it from the product’s specification.
How do white-box testing techniques work?
White-box techniques derive tests from internal structure, such as control flow. They can account for implementation paths when a specification is incomplete, vague, or out of date, but executing code paths does not establish that the software meets user needs. The CTFL v4.0 material highlights statement testing and branch testing. ASTQB’s presentation of the syllabus defines a branch as a transfer of control between nodes in a control-flow graph, whether conditional or unconditional.
Statement coverage
Statement coverage asks whether each executable statement has been run. Consider this illustrative pseudocode:
Do these 3 things before closing this tab:
1Scan for outdated or missing drivers - takes under a minute2Repair Windows errors before they cause bigger problems3Fix the driver behind crashes, sound loss and screen glitchesif (quantity > 0) {
charge(quantity);
}
writeAuditRecord();A test with quantity 2 executes the condition, the charge statement, and the audit statement. It therefore exercises every statement shown, but it does not test what happens when quantity is zero or negative.
Branch coverage
Branch coverage asks whether each possible control-flow outcome has been exercised. For the condition above, test quantity 2 to take the true branch and quantity 0 to take the false branch. The second test also checks whether the audit record is written when no charge occurs. Branch coverage makes the untested path visible, but neither statement nor branch coverage proves that the expected behavior is correct or that the requirements are complete.
Report the measure and denominator clearly: for example, “all executable statements in this component” or “both outcomes of this decision.” Coverage is evidence about which code structure ran, not a standalone quality score.
How do experience-based techniques find defects?
Experience-based testing uses the tester’s knowledge of the domain, product, technology, and prior failures to select probes. ISTQB CTFL v4.0 says: “Experience-based test techniques can detect defects that may be missed using the black-box test techniques and white-box test techniques.” These methods complement systematic techniques and are skill-dependent.
Error guessing
Error guessing turns knowledge of likely mistakes into focused tests. For the checkout example, a tester might probe a quantity that is blank, a repeated submit, an expired payment session, or an item whose stock changes during checkout—but only where those behaviors are relevant to the product. Use defect history and domain knowledge to choose the probes, then record the condition and observed result so the finding can become a repeatable test.
Exploratory testing
Exploratory testing combines learning, test design, execution, and evaluation. What the tester discovers guides the next action. Make it reproducible with a short charter, a timebox, notes, and follow-up cases. For example:
Rank #4
- Charter: Explore how checkout handles changes in stock while a customer is reviewing the order.
- Timebox: Set a defined session length appropriate to the team’s work; there is no universal duration.
- Observations: Record the starting data, actions, results, relevant environment details, and any unexpected behavior.
- Follow-up: Turn reproducible failures or important uncovered behavior into regression cases or clearer acceptance criteria.
Checklist-based testing
A checklist applies known risk prompts consistently without prescribing every click in advance. A checkout checklist might ask whether empty and malformed values are handled, totals are recalculated after edits, and a failed payment leaves the order in the specified state. Tailor prompts to actual requirements and past defects; a checklist is a memory aid, not proof that every behavior has been covered.
How can collaboration make requirements testable?
Some test design work belongs before implementation. Collaborative user-story writing, explicit acceptance criteria, and acceptance test-driven development (ATDD) help the people defining, building, and testing a feature agree on examples and expected behavior. CTFL v4.0 covers collaboration-based approaches alongside the core technique families. Clear conditions and examples give later black-box tests a more dependable basis.
For a checkout rule, agree on what “in stock” means at the moment of purchase, how a failed payment is reported, and whether an order can be retried. Express those decisions as examples or acceptance conditions. If stakeholders cannot agree on an expected result, adding more test cases will not resolve the underlying ambiguity.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.Where do techniques fit in the testing lifecycle?
Distinguish test levels from test types
Test levels group activities by the scope of software under test. CTFL lists five: component, component-integration, system, system-integration, and acceptance. Test types relate to quality characteristics or testing approaches; CTFL addresses functional, non-functional, black-box, and white-box types. Most types can be performed at different levels. For example, a functional test can apply to a component or a whole system, and black-box or white-box techniques can inform tests at different levels.
Choose the level according to the boundary you need to check, then choose a test type and technique suited to the risk. Do not assume that a technique belongs exclusively to one level.
After a code change: confirmation and regression
After a defect fix or enhancement, confirmation testing checks whether the specific fix works. Regression testing checks whether the change adversely affected other areas. ISTQB’s CTFL material says post-change testing should include both. As practical guidance, choose regression scope according to the change, affected dependencies, and risk; that is a planning judgment, not a universal fixed rule.
Quick wins for a faster PC:
Fix the driver behind crashes, sound loss and screen glitchesFind Drivers →Repair Windows errors before they cause bigger problemsFix Now →Best Value
How can screenshots support testing?
Screenshots can be useful visual evidence in interface checks, defect reports, or manual test records, but a screenshot alone does not verify behavior, state, or accessibility. Capture the relevant state consistently and retain the steps and expected result alongside the image. For automated UI testing, use assertions and appropriate visual comparison in addition to any captured artifact.
If you need a clean website capture as test evidence or a visual reference, ScreenshotNeo is a website screenshot API and MCP server. It is a supporting capture tool, not a substitute for designing and executing the test cases above.
Or skip the browser setup
One GET request can return a screenshot or PDF. This cURL example captures a page as WebP; replace the URL with the page you are permitted to capture. See the ScreenshotNeo API documentation for request options.
curl -G "https://api.screenshotneo.com/v1/shot" -d access_key=YOUR_API_KEY --data-urlencode url=https://stripe.com -o shot.webp- Cookie and consent banners are accepted before capture, and more than 60 known consent platforms, newsletter popups, and chat widgets can be removed; each step can be turned off.
- Bot checks or CAPTCHAs, blank pages, timeouts, failed loads, and cache hits are not billed. Response headers report the page verdict and billing status.
- An MCP server provides
take_screenshot,get_page_info, andcapture_pdftools for Claude, Cursor, and other MCP clients. - The Free plan includes 1,000 screenshots per month with no card; paid plans start at $5 for 3,000 screenshots. Every feature is on every plan.
Sign up for ScreenshotNeo’s free plan to start with 1,000 screenshots a month and no card.
Frequently Asked Questions
How many test cases should I create for one feature?
There is no universal target. Create enough cases to cover the relevant partitions, boundaries, rules, transitions, and structural paths for the risk and scope in question; document what remains outside the chosen coverage.
Does 100% statement or branch coverage prove a feature is correct?
No. Coverage shows which measured statements or branches executed. It does not prove that the expected behavior matches user needs, that every requirement was tested, or that unmodeled risks are absent.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
