Depressed patients who showed decreases in resting corrugator muscle tension levels, as measured electromyographically, also showed improvement in clinical symptoms, as assessed by the Hamilton Psychiatric Rating Scale over a 2-week period. Patients showing good clinical improvement after 2 weeks had resting pretreatment corrugator levels higher than those showing little clinical improvement, suggesting that corrugator activity might also serve as a forecaster of subsequent change. These findings support Darwin's 1872 hypothesis concerning the role of the "grief" muscle in depression. Furthermore, psychophysiological recording of patterns of facial muscle activity may have value in the clinical assessment of depression and treatment effects.