BACKGROUND: Although Mendelian randomization (MR) studies have been conducted on the causal relationship of testosterone on colorectal cancer (CRC), the result remains controversial. We aimed to explore the genetically determined relationships between total testosterone (TT) and bioavailable testosterone (BT) with CRC using a larger sample size and more stringent methods to exclude confounding factors. METHODS: Based on genome wide association studies (GWAS) data of TT, BT and CRC, we utilized bidirectional two-sample MR methods to analyze their interrelationships. Causal relationship analysis was conducted using inverse variance weighting (IVW), MR-Egger, weighted median, simple mode and weighted mode. Sensitivity analyses were performed to examine the stability of the causal relationships. RESULTS: The bidirectional MR analysis revealed one standard deviation (SD) increase in genetically predicted BT increased the risk of CRC [IVW: odds ratio (OR) =1.834, 95% confidence interval (CI): 1.121-3.001, P=0.02] and there was no causal relationship of CRC on BT. There was no causal relationship between CRC and TT. CONCLUSIONS: The findings of this study revealed a causal effect of BT on the risk of CRC, and CRC may not affect BT levels. Additionally, there was no causal relationship found between CRC and TT. Our results enhance the understanding of the real causal relationship between testosterone and CRC.