Perkins DO, Loohuis LO, Barbee J, et al. - Given that in persons who meet research criteria for a high-risk syndrome, the 2-year risk of psychosis is about 15%−25%, researchers here investigated persons at high risk to determine polygenic risk score (PRS) prediction of subsequent psychosis in these individuals. In addition, they assessed the consequence of the addition of the PRS to a previously validated psychosis risk calculator. Up to 2 years of follow-up was performed in persons meeting research criteria for psychosis high risk (N = 764) and unaffected individuals (N = 279). The latest schizophrenia and bipolar genome-wide association studies form the bases for the PRS. Stressful life events, trauma, disordered thought content, verbal learning, information processing speed, and family history of psychosis were the included variables in the psychosis risk calculator. Findings support the value of PRS in discriminating psychosis converters from nonconverters. Among Europeans, higher PRS was reported in the psychosis converter group vs both the nonconverter and unaffected groups, however, the nonconverter group was comparable to the unaffected group. Among non-Europeans, the converters and nonconverters did not show a significant difference, however, relative to unaffected individuals, converters showed significantly higher PRS, and nonconverters and unaffected individuals did not differ in PRS. When PRS is added to a psychosis risk calculator, modest improvement in individualized psychosis risk prediction was noted. The analysis suggest a promising value of schizophrenia PRS in improving risk prediction in persons at high risk for psychosis, although its poor performance in persons of non-European ancestry limits its potential utility.