Scale-up from lab to production is always problematic for the development of pharmaceuticals. In granulation, an optimal formulation of binder solution determined in a lab scale is often different than that in a production scale. A new mathematical procedure to solve this scale-up problem is assessed. Granules were prepared in the two manufacturing scales (2- and 5-kg scale) by using a high-speed mixer granulator. In the manufacturing process, the binder solution plays an essential role in the formation of granules with desired physical properties, in close conformity with the manufacturing scale. A computerized optimizing technique based on a response surface methodology was developed to study the scale-up problem in the manufacturing of granules. For this purpose, a new mathematical function was introduced for the first time, which is namely an integrated optimization function. A universal optimal formulation unaffected by manufacturing scale could be obtained by minimizing the integrated optimization function. Predicted values such as yield, mean granule size, and uniformity of granule size agreed well with experimental ones on both scales. Furthermore, the optimized characteristics measured at the production scale coincided well with those obtained at laboratory scale, suggesting that this approach could be very useful in minimizing scale-up problems.