Objective: It is likely that cognitive deficits are vulnerability markers for developing schizophrenia, as these deficits are already well-established findings in first-episode psychosis. Studies at-risk adolescents and young adults are likely to provide information about cognitive deficits that predate the onset of the illness. Method: We conducted meta-analyses of studies comparing familial-high risk (FHR) or ultra-high risk (UHR; n = 2113) and healthy controls (n = 1748) in youth studies in which the mean age was between 15 and 29. Results: Compared with controls, high risk subjects were impaired in each domain in both UHR (d = 0.34-0.71) and FHR (d = 0.24-0.81). Heterogeneity of effect sizes across studies was modest, increasing confidence to the findings of the current meta-analysis (I2 = 0-0.18%). In both risk paradigms, co-occurrence of genetic risk with attenuated symptoms was associated with more severe cognitive dysfunction. In UHR, later transition to psychosis was associated with more severe cognitive deficits in all domains (d = 0.31-0.49) except sustained attention. However, cognitive impairment has a limited capacity to predict the outcome of high-risk patients. Conclusion: Cognitive deficits are already evident in adolescents and young adults who have familial or clinical risk for psychosis. Longitudinal developmental studies are important to reveal timing and trajectory of emergence of such deficits.