Based on data drawn from a random-digit-dial probability sample of adults in a major American metropolitan area, this study supports the perspective that self-efficacy is domain specific and that outcome efficacy is distinct from self-efficacy. Data were collected by telephone interviewers which were administered by telephone interviews to 525 respondents with a 64.1% completion rate. Sample demographic summary statistics closely approximated population parameters. Five orthogonal factors emerged from analysis of self-efficacy and outcome efficacy items. The five factors represented self-efficacy with regard to nutrition, medical care, exercise, and with a set of neutral control items relating to political behavior, and with the outcome efficacy items for each behavioral domain. Hypotheses relating scores for each factor with a number of behavioral indicators were tested. Only four of 125 correlations failed to support hypothesized relationships, lending evidence for the discriminant validity of the self-efficacy dimensions.