In the self-regulation of anger, it is unclear whether the effects are due to self-monitoring or actual self-intervention. This is especially important in the light of the reactivity often attributed to self-monitoring. The present study compared both procedures within a reversal design involving college students with moderate levels of anger. Intervention consisted of behavioral contracting and relaxation as well as cognitive strategies of reappraisal and imagery. Measures included anger frequency, duration, and intensity, all of which were obtained in naturalistic settings. Results revealed significant and dramatic reductions in anger frequency and duration during self-intervention but no such improvements with self-monitoring apart from a slight reduction in anger intensity. The return to baseline self-monitoring was associated with no relapse in anger but a maintenance of treatment gains. Secondary findings of interest were a correlation between anger intensity and anger duration and a predominance of anger directed at persons rather than situations. In general, the outcome of this study adds strength to the inference that cognitive-behavioral self-interventions are effective for anger and that this effectiveness goes beyond mere reactivity. The effects are most pronounced for anger frequency and duration and this offers clues for future focal points in the practice of anger management.