ABSTRACT: The initiators caspase-9 (CASP9) and caspase-10 (CASP10) are two key controllers of apoptosis and play important roles in carcinogenesis. This study aims to explore the association between CASPs gene polymorphisms and colorectal cancer (CRC) susceptibility in a population-based study. A two-stage designed population-based case-control study was carried out, including a testing set with 300 cases and 296 controls and a validation set with 206 cases and 845 controls. A total of eight tag selected single nucleotide polymorphisms (SNPs) in CASP9 and CASP10 were chosen based on HapMap and the National Center of Biotechnology Information (NCBI) datasets and genotyped by restriction fragment length polymorphism (RFLP) assay. Multivariate logistic regression models were applied to evaluate the association of SNPs with CRC risk. In the first stage, from eight tag SNPs, three polymorphisms rs4646077 (odds ratio (OR)(AA+AG): 0.654, 95% confidence interval (CI): 0.406-1.055; P=0.082), rs4233532 (OR(CC): 1.667, 95% CI: 0.967-2.876; OR(CT): 1.435, 95% CI: 0.998-2.063; P=0.077), and rs2881930 (OR(CC): 0.263, 95% CI: 0.095-0.728, P=0.036) showed possible association with CRC risk. However, none of the three SNPs, rs4646077 (OR(AA+AG): 1.233, 95% CI: 0.903-1.683), rs4233532 (OR(CC): 0.892, 95% CI: 0.640-1.243; OR(CT): 1.134, 95% CI: 0.897-1.433), and rs2881930 (OR(CC): 1.096, 95% CI: 0.620-1.938; OR(CT): 1.009, 95% CI: 0.801-1.271), remained significant with CRC risk in the validation set, even after stratification for different tumor locations (colon or rectum). In addition, never tea drinking was associated with a significantly increased risk of CRC in testing set together with validation set (OR: 1.755, 95% CI: 1.319-2.334). Our results found that polymorphisms of CASP9 and CASP10 genes may not contribute to CRC risk in Chinese population and thereby the large-scale case-control studies might be in consideration. In addition, tea drinking was a protective factor for CRC.