Unlike negative risk-taking, positive risk-taking is thought to support young people’s desire for exploration and novelty in socially acceptable and developmentally appropriate ways. Although positive risk-taking is hypothesized to reflect a more optimal pattern of decision-making in the real world, scholars have yet to establish this behaviorally in the lab. In the present study, we assessed negative (e.g., substance use, shoplifting, getting into fights) and positive risks (e.g., starting new friendships, enrolling in a new class, standing up for one’s beliefs) among a sample (n = 171) of people ages 14−30. We then examined whether and how self-reported positive and negative risk-taking are predicted by risk-taking, performance, learning, and sensitivity to negative feedback on the Balloon Emotional Learning Task (BELT). While positive risk-taking was expected to demonstrate a more optimal decision-making pattern, the findings did not fully support this hypothesis. The propensity for negative risk-taking was associated with greater risk-taking and lower sensitivity to negative feedback on the BELT, whereas the associations between these variables and positive risk-taking were less evident. The propensity for positive risk-taking was associated with poorer task performance in a condition that does not allow for learning. The propensity for negative risk-taking was associated with poorer or more optimal task performance, depending on whether the probability of negative consequences was high or low. Together, findings shed light on the complexity of risk-taking behaviors and open avenues for future research examining the shared and unique factors linked to positive and negative risk-taking in the lab.