Two studies were carried out to determine and compare the effects of several sources of variation on the measurement of corneal thickness using the standard optical pachymeter and three ultrasonic pachymeters. Sources of variation included: intra- and inter-session variation, inter-observer variation, left/right eye variation, and variations due to alternate settings of ultrasonic sound frequencies. It was found that the optical pachymeter had a) two to three times as much intra-session variation as that of the ultrasound pachymeters, b) significant inter-observer variation (P = 0.015), and c) significant differences between left and right eye thickness determinations (P less than 0.005). On the other hand, ultrasonic pachymeters demonstrated a) high reproducibility, b) no inter-observer variation, and c) no left/right eye variation. These results have implications for the use of pachymetry in measuring corneal thickness for radial keratotomy and other refractive surgery.