Software engineering requires rigorous testing to guarantee the product's quality. Semantic testing of functional correctness is challenged by nondeterminism in behavior, which makes testers difficult to write and reason about. This thesis presents a language-based technique for testing interactive systems. I propose a theory for specifying and validating nondeterministic behaviors, with guaranteed soundness and correctness. I then apply the theory to testing practices, and show how to derive specifications into interactive tester programs. I also introduce a language design for producing test inputs that can effectively detect and reproduce invalid behaviors. I evaluate the methodology by specifying and testing real-world systems such as web servers and file synchronizers, demonstrating the derived testers' ability to find disagreements between the specification and the implementation.