Computer optimization technique was applied to the simultaneous optimization of wet granulation process by a high-speed mixer granulator. Four pharmaceutical properties, including yield, drug content uniformity, geometrical mean diameter of granules, and uniformity of granule size, were selected to evaluate the quality of the granules. In particular, dependence of drug content uniformity on granule size was investigated using two model drugs, ascorbic acid and ethenzamide. An appreciable dependence of ascorbic acid content on granule size was not observed in model formulations. On the other hand, ethenzamide was contained more in small-size granules, and its content was decreased with an increase in amounts of hydroxypropyl cellulose (HPC-L; used as a binder) and binder solution. These observations suggested that drug content uniformity is influenced not only by drug solubility in the binder solution, but also by the use of HPC-L. A simultaneous optimal point incorporating four pharmaceutical properties was obtained using the generalized distance function. The experimental values of the four response variables obtained in newly prepared granules were found to correspond well with the predicted values of both granules containing ascorbic acid and ethenzamide. These results suggested that computer optimization would benefit the wet granulation process even if drug content segregation was involved in the process. Further, data obtained from computer optimization, in particular the contour diagram, will be valuable in the process validation.