It has been suggested that self-reported, common, non-clinical psychotic experiences may represent the transitory developmental expression of distributed genetic risk for psychosis. In a sample of female MZ (176 pairs) and DZ twins (113 pairs), cross-twin, cross-trait analyses were conducted to investigate the association between repeated continuous measures of self-reported psychotic experiences (PE—three measures over 18 months), assessed with the CAPE, in one twin and clinical interview categorical measures of psychotic symptoms (PS), assessed with SCID-I, in the other twin. The results showed that in MZ but not DZ pairs (interaction: χ2 = 7.9, df = 1, P = 0.005), the cross-twin association between PE and PS was large and significant (standardized effect size: 0.26, 95% CI: 0.10–0.42) and of similar magnitude as the within-twin PE–PS association (standardized effect size: 0.28, 95% CI: 0.10–0.45), demonstrating both PE validity and genetic effects. In addition, the cross-twin association between PE and PS was significantly larger (interaction: χ2 = 20.3, df = 1, P < 0.0001) for younger MZ twins (standardized effect size: 0.67, 95% CI: 0.44–0.90) than older MZ twins (standardized effect size: −0.05, 95% CI: −0.26 to 0.16), demonstrating developmental effects. This study indicates that self-reported psychotic experiences in the general population may represent the developmental expression of population genetic risk for psychosis. © 2009 Wiley-Liss, Inc.