The aim of this study was to examine the effect of repeated testing on prepulse inhibition (PPI) and habituation of the startle reflex. Fifteen healthy control subjects (eight males, mean age 30 years; seven females, mean age 29 years) were tested on three occasions across the same day separated by a minimum of 2 h. An acoustic probe of 40-msec bursts of 116 dB(A) white noise over a continuous background noise of 70 dB(A) was presented binaurally through headphones and the eye-blink component of the startle response was measured taking electromyographic recordings from the right orbicularis oculi. The test session was identical at each time point and consisted of two blocks of 12 randomly mixed trials of four pulse-alone, four 60-msec prepulse and four 120-msec prepulse trials enclosed by two blocks each of six pulse-alone trials. There was huge variation in individual response magnitude that was independent of subsequent PPI in both women and men. Women showed greater PPI in the second half of sessions with the 120-msec prepulse only; but PPI was not altered significantly in either group between sessions across the day. In general, there was good test-retest reliability of PPI especially within trial type. Normal reflex habituation occurred across sessions and this effect was preserved in sessions across the day. Latency of response was significantly reduced in a session by the 60-msec trial type compared to the 120-msec trial type, as previously reported. Our results suggest that measures of PPI and habituation of the startle response are appropriate and reliable for a within-subject, test-retest design.