This article provides recommendations for incorporating conceptual models of health behavior change into research conducted in emergency care settings. The authors drafted a set of preliminary recommendations, which were reviewed and discussed by a panel of experienced investigators attending the 2009 Academic Emergency Medicine consensus conference. The original recommendations were expanded and refined based on their input. This article reports the final recommendations. Three recommendations were made: 1) research conducted in emergency care settings that focuses on health behaviors should be grounded in formal conceptual models, 2) investigators should clearly operationalize their outcomes of interest, and 3) expected relations between theoretical constructs and outcomes should be made explicit prior to initiating a study. A priori hypothesis generation grounded in conceptual models of health behavior, followed by empirical validation of these hypotheses, is needed to improve preventive and public health-related interventions in emergency care settings.