We examined heritability of error rate on the antisaccade task among female twin youths. This task appears to be sensitive to prefrontal functioning, providing a measure of individual differences in inhibitory control associated with genetic risk for schizophrenia. The sample consisted of 674 11-year-olds and 616 17-year-olds, comprising the two cohorts of female twins from the Minnesota Twin Family Study, a population-based investigation of substance abuse and related psychopathology. We used biometric model-fitting methods to determine the relative magnitude of genetic and environmental influences on performance. In both age cohorts, the best fitting model contained additive genes and nonshared environment. Despite substantial age-related differences in mean performance levels (effect size = .81), additive genes accounted for greater than half the variance in performance in both age cohorts. These results are consistent with the hypothesis that antisaccade error rate might serve as an endophenotype for behavior disorders reflecting frontal lobe dysfunction or problems with inhibitory control.