This article reports two studies designed to test a theory of planned behavior-based model of blood donation. In Study 1 (n= 136), self-efficacy and self-identity accounted for unique variance in behavioral intention. Study 2 (n= 172) extended Study 1: self-efficacy, perceived control over behavior (PCB), self-identity, and moral norm were all independently predictive of intention; behavioral intention predicted a proxy measure of behavioral enaction. Both studies provided evidence to support a distinction between self-efficacy and PCB, and for the inclusion of self-identity and moral norm into the model. Belief-based measures discriminated intenders from nonintenders, and beliefs that accounted for unique variance in self-efficacy and PCB were identified. The findings are discussed in relation to using models such as the theory of planned behavior to intervene in social and health behaviors.