In a placebo-controlled, randomized, double-blind trial involving outpatients with mild to moderately severe depression, an extract of St. John's wort (Hypericum), LI 160, a herbal antidepressant was tested for efficacy and tolerability, as well as for possible negative effects on cognitive performance. The responder rate to treatment with the extract was 66.6% as compared with only 26.7% with placebo. The treatment was very well tolerated; only in two patients did transient minor side effects occur under LI 160. No impairment of cognitive performance was observed: during the trial, Hypericum did not lead to any impairment of attention, concentration or reaction.