The areas of 50 randomly chosen skin-test reactions were measured with a graphics tablet interfaced to a microcomputer. The computer-determined areas were compared with the sums and products of orthogonal diameters, the maximum diameter, and the actual areas of the reactions. The areas measured by the computer were highly correlated with the actual areas (r = 0.999). Both the sums and products of orthogonal diameters were also highly correlated with the actual areas (r = 0.953 and r = 0.988, respectively) and added little to the correlation of the maximum diameter (r = 0.923). The average coefficient of variation for repeated computer area measurements was 5.3%. Area and maximum diameter measurements were then compared in a study designed to evaluate the variation of prick tests applied by different persons. Area measurements revealed significant differences between the testers, which were not found when maximum diameter measurements were compared. Computerized area determination is a rapid, accurate means for quantitating skin-test reactions. The precision offered by this method is of potential importance for studies involving quantitative skin testing.