A group of eighteen adults with refractory epileptic seizures were given psychological treatment in a two-phase experimental group study. In phase one, the experimental phase, the patients were divided into three groups--contingent relaxation (CR), attention control (ATC) treatment, and a no-treatment (NT) control group--with the purpose of investigating the effects of a learning-based contingent relaxation program compared with the effects of professional attention alone when superimposed on a regular medical treatment program. The design of the experimental phase was comprised of a 10-week baseline, 6-week intervention, and 10-week follow-up. Results of this phase at the end of follow-up showed a significant reduction only for those patients receiving the CR treatment. In the nonexperimental phase, the two control groups also received the CR treatment for a 6-week period, and subsequent seizure frequency measures for all three groups were analyzed after 10-week and 30-week follow-up periods. Results of this phase showed a significant reduction in seizure frequency for all three groups after receiving the CR treatment. Effects of the CR treatment were maintained at a 30-week follow-up. The results indicate that the CR treatment program may be of substantial help to adults whose seizures are resistant to conventional drug therapy.