In this work we investigated the effects of the drum rotation speed, liquid-phase ratio, granulation time, and drum hold-up on the single super-phosphate granulation performance in a rotary drum. Regression techniques were used to quantify the effects of those variables on the granulation efficiency and product properties. Equations to predict these responses as functions of the studied variables were proposed. Sequential Quadratic Programming (SQP) algorithm was used for the optimization study. The results showed that the levels of the process variables that led to the highest efficiency were not the same for the highest granule hardness. From the results of this study it is possible to identify conditions that improve the granulation efficiency and the physical properties of the final product. (C) 2017 Elsevier B.V. All rights reserved.