Black-box testing treats software as an opaque component: tests use public interfaces, observe produced results, and avoid relying on internal implementation knowledge. Its purpose is to reduce project risk, not to guarantee bug-free code. section 15.1 section 15.3
Use the boundary to distinguish black-box testing from related approaches:
- Black-box tests remain unaware of implementation details; grey-box and white-box tests use partial or complete implementation knowledge. section 15.1
- Functional tests may be black-box or grey-box, while integration tests exercise connected application parts and subsystem boundaries. section 15.1
- Black-box testing alone cannot compensate for an architecture with many external services and uncontrolled side effects. section 15.2
Testing through an interface
A test calls a component through its public interface, supplies an input, and checks the returned result or externally visible behavior. The test remains valid without knowing how the component computes that result. section 15.1






