We determined the magnitude of variability in optic disc topographical parameters on digital analysis of the optic disc using the IS 2000. The variability introduced by the system, the observer, the observer and patient, and by clinically different types of discs was assessed in the measurement of the vertical cup-to-disc ratio, horizontal cup-to-disc ratio, cup area-to-disc area, cup volume, neuroretinal rim area, and neuroretinal rim area-to-disc area. The system itself accounted for no variability. The variability introduced by one observer for the parameters ranged from 1% to 7%, and by one observer and patient from 1% to 28%. The variation among five observers ranged from 1% to 55%. Direct image acquisition (using video cameras) gave results that were no different from those obtained by digitizing the slides. Contrast-enhancement techniques did not decrease observer variability. A change in the flash intensity level at which optic disc images were acquired from 94.5 to 15.0 Watt-seconds introduced a variability of 3% to 21%. These results are less variable than those obtained on clinical observation and comparable to those of the Rodenstock image analyzer in evaluating these aspects of optic disc topography.