A longitudinal panel study examined the effects of actual use of diazepam (Valium) on subjective reports of life quality, affect, performance, stress, social support, control, coping and other variables related to mental health. Standardized interviews were conducted with 675 persons from the Detroit Metropolitan Area. Based on prescription records, diazepam users and nonusers were selected to represent a variety of sociodemographic characteristics rather than to be a completely random sample. Significant others in work and in personal life were also interviewed. Four interviews took place, one approximately every 6 weeks. Testing for social effects was conducted by within- and across-person analyses of 367 respondents who reported taking the medication at some time during the study and by comparisons with 308 respondents who did not report taking Valium. Users of Valium tended to take less Valium than prescribed. They also reported consuming less alcohol when using Valium than at other times and less than non-Valium users. Although there was a modest, positive cross-sectional relation between Valium use and distress, numerous multivariate analyses controlling for levels of stress and health indicated no notable effects of Valium use on any of the social or psychological indicators, including anxiety. Several interpretations of the results are examined including the possibility that the effects of Valium use were short-lived rather than long-term and that Valium may have been taken in anticipation of anxiety rather than after its occurrence.