A testing strategy in which the smallest unit of testable code is isolated from the rest of the software and tested to determine if it functions properly.