Salivary testosterone concentrations were measured in male and female members of four heterosexual couples on a total of 11 evenings before and after sexual intercourse and 11 evenings on which there was no intercourse. Testosterone increased across the evening when there was intercourse and decreased when there was none. The pattern was the same for males and females. Early evening measured did not differ on the two kinds of days, suggesting that sexual activity affects testosterone more than initial testosterone affects sexual activity.