The suitability of single subject designs for research on coping skills training is discussed. Such training programs teach the individual an active skill for dealing with anxiety-provoking situations. Several research strategies are reviewed, with a discussion of their major assumptions and their potential for testing for generalized effects: (1) reversal, or withdrawal of the treatment condition; (2) simultaneous treatment, in which different interventions are compared; (3) changing criterion, in which stepwise change in behavior is associated with each phase of treatment; and (4) multiple baseline, in which baseline data are first stabilized on several subjects; the behavior of the experimental subject does not change until he or she is exposed to the treatment condition. It is concluded that the multiple baseline design is the most appropriate strategy for single subject coping skills training. (Author/GDC)