Graham (1975) demonstrated that a weak prestimulus could effectively inhibit or facilitate the eyeblink component of the startle reflex in humans, depending on the temporal duration of the prestimulus. This study had three goals: 1) to replicate the findings of Graham, 2) to establish the reliability of this phenomenon by a test‐retest comparison, and 3) to compare the eyeblink reflex response of normal subjects with schizophrenic subjects. Seven prestimulus durations of continuous tone (from 0 to 2000 msec) were presented to 20 normal subjects and the results confirmed that maximal inhibition of eyeblink amplitude occurred in the 120 msec prestimulus condition. Increased amplitude occurred nonsignificantly when the prestimulus lasted for 2000 msec. On retest, 14 normal subjects showed a significant degree of reliability. When 20 normal subjects were compared to 12 schizophrenic subjects, significant differences in eyeblink response were found for blink amplitude and latency in the 60 msec prestimulus condition. This change is consistent with information processing “overload” theories of sensory overstimulation in schizophrenia. The blink reflex is a rather stable phenomenon and is probably altered in schizophrenia and/or by antipsychotic medication.