To evaluate the effect of cognitive behavioral intervention on chronic fatigue syndrome (CFS), we studied three patient groups: a CFS-treatment group (n = 22), a primary depression-treatment group (n = 20), and a no-treatment control group of subjects with CFS (n = 22). For the CFS-treatment group, a trend toward reduced depression-symptom scores was noted, but there were no significant changes in stress-related symptoms or fatigue severity. For the most depressed treated subjects with CFS, significant score reductions were observed in measures of depression, stress, fatigue severity, and fatigue-related thinking. In the depression group, significant reductions in depression, stress, and fatigue severity scores were found. No significant changes in any measure were observed in the CFS control group. A new fatigue-related cognitions scale, developed to assess cognitive and emotional reactions to fatigue, showed a significant reduction in such reactions in the CFS-treatment group, a finding suggesting that depression in this group was mediated by maladaptive thinking. The results suggest that a subset of CFS patients with cognition-related depressive symptomatology may respond to short-term behavioral intervention.