The Zeiss Scheimpflug Cataract Video Camera was designed to photograph, store and analyze cataracts in a semi-automated fashion for cross-sectional and longitudinal studies. We conducted a reproducibility study of this system. Twenty-four normal and 61 cataractous eyes were photographed twice by each of two of the authors in the 90 degree meridian and microdensitometry was performed on each of the stored images. Reproducibility was then determined using the intraclass correlation coefficient to determine whether or not the differences encountered were due to variability in the system or due to actual differences among the images. The intraclass correlation in the lens nucleus was 0.995 with 95% confidence limits of .992-.996. Therefore, reproducibility was 99.5%. In the anterior cortex, intraclass correlation was .941 with 95% confidence limits of .919-.959. In the posterior cortex intraclass correlation was .905 with 95% confidence limits of .870-.932. Reproducibility with this instrument was therefore excellent and with certain limitations, this may be a useful instrument in monitoring lens changes in certain diseases and the effects of anti-cataract agents.