Irinotecan has been used in the first-line treatment of metastatic colorectal cancer. However, no clear predictive marker of irinotecan efficacy has been identified. It is controversial whether the response to irinotecan could be predicted by the expression level of topoisomerase-I, a direct target of irinotecan. The present study aimed to identify a feasible predictive marker of irinotecan efficacy. We hypothesized that the efficacy of SN38 (an active metabolite of irinotecan) is related to the cell proliferation and the phosphorylation status of RB in colorectal cancer cells. Indeed, the IC50 of SN38 was positively correlated with the doubling time of each cell line (R2=0.9315). Moreover, the phosphorylation level of RB was related to SN38 sensitivity. Consistent with the in vitro data, colorectal cancer tissues of irinotecan responders showed a significantly higher rate of phosphorylated RB (serine 780) expression using immunohistochemistry (P=0.0006), although a generally used proliferative marker, Ki-67, showed no significance. Finally, we investigated whether the phosphorylation of RB plays a crucial role in the efficacy of irinotecan. To suppress the expression of phosphorylated RB, we performed the knockdown of CDKs, which are known to phosphorylate RB. Intriguingly, the knockdown of both CDK4 and CDK6, but not CDK2, allowed RB to become the most hypophosphorylated form and converted the SN38-sensitive cells to a resistant state. Taking together the above findings from in vitro and clinical research, the immunohistochemistry of phosphorylated RB protein might be feasible to predict the irinotecan efficacy of colorectal cancer in clinical practice.