Within- vs across-Ss procedures for predicting behavior from attitudes were contrasted. Each procedure required a comparison among attitudes to generate a prediction; the comparison was either among the same attitudes held by different people (across Ss) or among different attitudes held by the same person (within Ss). It was hypothesized that the within-S model would provide a more adequate explanation of behavior from attitudinal constructs and, hence, more accurate prediction of behavior from attitudes than the across-Ss model. 349 married couples were administered a questionnaire containing measures of 3 attitudinal components—affect, cognition, and conation—toward each of 4 contraceptive methods—oral contraceptives, intrauterine device, diaphragm, and condom. Contraceptive behavior was assessed 1 yr later. In support of the hypothesis, the within-S predictions bore a significantly stronger relation to the behavioral criteria than did the across-Ss predictions. This effect was consistent for each of the 3 components of attitude, for the prediction of behavior and behavioral intention, for male and female Ss, and for a variety of contraceptive behaviors. In addition, both the within- and the across-Ss analyses demonstrated a clear rank ordering in the predictive validity of the 3 attitudinal components: Conation was a better predictor of behavior than was the affect, which, in turn, was better than cognition. (2 p ref) (PsycINFO Database Record (c) 2016 APA, all rights reserved)