A sequential risk-taking paradigm used to identify real-world risk takers invokes both learning and decision processes. This article expands the paradigm to a larger class of tasks with different stochastic environments and different learning requirements. Generalizing a Bayesian sequential risk-taking model to the larger set of tasks clarifies the roles of learning and decision making during sequential risky choice. Results show that respondents adapt their learning processes and associated mental representations of the task to the stochastic environment. Furthermore, their Bayesian learning processes are shown to interfere with the paradigm's identification of risky drug use, whereas the decision-making process facilitates its diagnosticity. Theoretical implications of the results in terms of both understanding risk-taking behavior and improving risk-taking assessment methods are discussed.Keywords: risk taking, learning, Bayesian, individual differences, cognitive psychometrics Learning and decision making are conceptually linked. Typically only after decision makers (DMs) make a decision do they observe or experience the precise outcome of that decision. For example, only after commuters select a traffic route do they determine its effectiveness, and only after athletes use a steroid do they learn about the precise properties it has on their body. These observations better inform DMs about the precise properties of their choice options and shape their next decision among the same or similar options. Despite this natural association between decision and learning processes, most decision-making theories fail to incorporate or explicate a learning component (e.g., Busemeyer & Townsend, 1993;González-Vallejo, 2002;Kahneman & Tversky, 1979). Yet, how DMs learn from experience has proven an important process in understanding risk-taking behavior. It can, for example, create an aversion toward risky alternatives in the gain domain and an attraction toward risky alternatives in the loss domain-a pattern typically attributed to how DMs evaluate outcomes (Denrell, 2007;March, 1996). The learning process can even produce the opposite pattern (Erev & Barron, 2005;Hertwig, Barron, Weber, & Erev, 2004;Weber, Shafir, & Blais, 2004).Applying theories of decision making to the Balloon Analogue Risk Task (BART; Lejuez et al., 2002) or to the Iowa Gambling Task (Bechara, Damasio, Damasio, & Anderson, 1994) also exposes the necessity of learning. Clinicians use these laboratorybased gambling tasks to study and identify people with specific clinical or neurological deficits. Cognitive models of these tasks reveal that decision and learning processes are necessary to account for choices made by both clinical and normal populations (Busemeyer & Stout, 2002;Wallsten, Pleskac, & Lejuez, 2005). Besides describing behavior during the tasks, the models also show how the populations differ on the underlying cognitive dimensions captured within the models. For example, during the BART, Wallsten et al. (2005) found that people who take un...