The development of effective prevention initiatives requires a detailed understanding of the characteristics and needs of the target audience. To properly identify at-risk individuals, it is crucial to clearly delineate risky from acceptable behavior. Whereas health behavior campaigns commonly use single conditions (e.g., lack of condom use) to identify high-risk cohorts, many risk behaviors are more complex and context dependent, rendering a single condition approach inadequate. Out-of-bounds skiing, an activity associated with voluntary exposure to avalanche hazard, is an example of such a multifaceted risk-taking activity. Using a dataset from an extensive online survey on out-of-bounds skiing, we present an innovative approach for identifying at-risk individuals in complex risk environments. Based on a risk management framework, we first examine risk-taking preferences of out-of-bounds skiers with respect to exposure and preparedness—the two main dimensions of risk management—separately. Our approach builds on existing person-centered research and uses Latent Class Analysis to assign survey participants to mutually exclusive behavioral classes on these two dimensions. Discrete Choice Experiments are introduced as a useful method for examining exposure preferences in the context of variable external conditions. The two class designations are then combined using a risk matrix to assign overall risk levels to each survey participant. The present approach complements existing person-centered prevention research on the antecedents of risk-taking by offering a process-oriented method for examining behavioral patterns with respect to the activity itself. Together, the two approaches can offer a much richer perspective for informing the design of effective prevention initiatives.