We compared event-related brain potentials (ERPs) and reaction time (RT) measures from untreated young (prepubertal and peripubertal; ages 9 to 14) and old (postpubertal; ages 15 to 20) Turner's syndrome (TS) subjects with those from normal age-matched controls. Comparisons among groups permitted the assessment of the relative roles of congenital and maturational brain alterations as possible bases of the cognitive deficits in TS. All subjects were presented with series of auditory stimuli, and they either counted one of the two stimuli or made rapid discriminative button presses to both. The results indicated that, whereas the ERPs in young TS females were essentially the same as those in their age-matched controls, the ERPs in old TS females more closely resembled those in both young groups than those in their age-matched controls. Specifically, a late (400 to 900 msec) frontal negative slow wave (Nc) in the old TS subjects failed to show the normal maturational course in which the amplitude and duration of this component steadily decreases with age. Except for slightly greater amounts of N1 amplitude in the young TS group, the latencies, amplitudes, and scalp distributions of the other ERP components (N1, P2, N2, P300) were all the same at all ages in the TS subjects and their controls. Behaviorally, both TS groups had longer RTs than their controls but, consistent with the ERP results, they did not make more errors. The ERP and RT results point to the existence of two kinds of abnormalities in TS females, each from a different mechanism: the ERP results suggest an age-dependent maturational defect while the RT results suggest an age-independent congenital defect.