This talk reviews some of my contributions on formal testing of timed and probabilistic systems, focusing on methodologies that allow their users to decide whether these systems are correct with respect to a formal specification. The consideration of time and probability complicates the definition of these frameworks since there is not an obvious way to define correctness. For example, in a specific situation it might be desirable that a system is as fast as possible while in a different application it might be required that the performance of the system is exactly equal to the one given by the specification. All the methodologies have as common assumption that the system under test is a black-box and that the specification is described as a timed and/or probabilistic extension of the finite state machines formalism.