To develop an efficient index for evaluating ulcerative colitis, we quantitatively assessed disease activity by multiple regression analysis. Thirty-two consecutive cases with ulcerative colitis were studied. These cases were divided by degree of severity into three groups: mild, moderate, and severe; this tripartite grouping was used as a dependent variable. As independent variables, 11 predictors were selected from among physical findings and laboratory data. By multiple regression analysis with stepwise deletion, six factors indispensable to the disease activity were selected: temperature, pulse rate, erythrocyte sedimentation rate, red blood cell count, alpha 2-globulin, and number of bowel movements per day. Disease activity is predicted by using the following equation: Y = -12.168 + 0.333X1 + 0.022X2 + 0.018X3 - 0.337X4 + 0.056X5 + 0.033X6, where X1, X2, X3, X4, X5, and X6 denote temperature, pulse rate, erythrocyte sedimentation rate, red blood cell count, alpha 2-globulin, and number of bowel movements per day, respectively. By applying this equation to each case, the range of activity indices is expressed as follows: mild, less than 1.52; moderate, between 1.52 and 2.44; and severe, more than 2.44.