Although societal need for behavioral science research is enormous, current research practices seem to be inefficient vehicles for producing knowledge that guides practical action. Many of our most popular theories provide little direct guidance for application. They focus on the development of models of the relationships among organismic events such as attitudes, self-efficacy expectations, and behavior, but pay little or no attention to the contextual influences on behavior. Such research is in keeping with a long-standing mechanistic tradition in psychology. We propose a version of contextualism as an alternative paradigm for the behavioral sciences. According to this paradigm, theories and research are evaluated in terms of their contribution to the prediction and influence of behavior. Basic research organized to pursue this goal has a direct bearing on how behavioral phenomena can be changed for practical purposes. Conversely, applied research contributes to basic understanding of the determinants of psychological phenomena.