This meta-analytic review aims to address the mixed findings in previous research by quantifying the associations between early-life stress and risk, time, and prosocial preferences, and testing the boundary conditions of these associations. We meta-analyze 123 articles reporting 867 effect sizes among 199,019 adults to test different predictions from a life history perspective, a sensitization perspective, and an uncertainty management perspective about how early-life stress is associated with risk, time, and prosocial preferences. First, we find relatively small effect sizes indicating that early-life stress is associated with greater risk taking (r = .123), more present orientation (r = .126), and less prosociality (r = -.085), and its positive association with present orientation is stronger in currently stressful situations. Second, these observed associations do not vary significantly for harshness and unpredictability dimensions of early-life stress. Notably, moderation analyses across different types of preference measures only reveal an overall pattern of associations of early-life stress with self report measures of risk, time, and prosocial preferences. By contrast, early-life stress is not significantly associated with risk preference or prosocial preference measured with hypothetical choice tasks or laboratory behavior tasks. Taken together, although the overall pattern of results supports a life history perspective, a cautious interpretation is warranted by the variation in the results across different preference measures and potential publication bias in the results. More pre-registered studies are needed to test the extent to which preferences measured with arbitrary laboratory-based tasks capture real-world behaviors and to increase the ecological validity of laboratory-based measures.