AbstractResponse surface methodology was used with four factors to screen for the best starch and optimize the use of chemicals in order to maximize precipitated calcium carbonate (PCC) filler retention in a peroxide-bleached TMP suspension. Three commercial starches were used in conjunction with colloidal silica and flocculant. The PCC loading level and the interactions between PCC level, starch, flocculant, and silica were investigated, and empirical models were constructed. The empirical process models were then employed to predict the retention and drainage. It was found that medium-charged cationic starch (S858) gave the highest total and filler retention, whereas high-charged cationic starch (S880) resulted in the best drainage. The ash content of the handsheet can be pushed up to 40% using the retention system with medium (S858) and high (S880) charged cationic starch. The high-charged cationic starch (S880) gave stronger paper, probably because of its higher affinity with the fiber and fines.