Background: Growth hormone (GH) may be associated with the development of colorectal tumors directly and/or indi- rectly via an increased plasma level of insulin-like growth factor-I (IGF-I), which has been associated with colorectal cancer risk. Because a T-to-A polymorphism in the human GH1 gene at position 1663 is putatively associated with lower levels of GH and IGF-I, we investigated the relationship of this polymorphism to the risk of colorectal neoplasia. Meth- ods: We analyzed data from two case-control studies con- ducted in Hawaii: a population-based study of 535 case pa- tients with colorectal adenocarcinoma and 650 control subjects and a sigmoidoscopy screening-based study with 139 case patients with adenoma and 202 control subjects. All subjects were tested for the GH1 polymorphism. Logistic regression was used to adjust for known risk factors. Plasma IGF-I and IGF binding protein-3 (IGFBP-3) levels were measured in a subset of 293 subjects in the adenoma study (135 case patients and 158 control subjects). All statistical tests were two-sided. Results: Adjusted odds ratios (ORs) for colorectal cancer associated with T/T, T/A, and A/A genotypes were 1.00, 0.75 (95% confidence interval (CI) = 0.58 to 0.99), and 0.62 (95% CI = 0.43 to 0.90), respectively (Ptrend = .006). Adjusted ORs for adenoma were 1.00, 0.76 (95% CI = 0.46 to 1.24), and 0.62 (95% CI = 0.31 to 1.22), respectively (Ptrend = .17). Data from both studies consis- tently showed that the A allele was associated with a lower risk of colorectal neoplasia than the T allele, although the association with adenoma was not statistically significant. These associations were consistently suggested in Caucasians and Native Hawaiians but not in Japanese. The ratio of plasma IGF-I/IGFBP-3 was lower in individuals with the A allele than in individuals with the T allele (P = .01). Con- clusion: The human T1663A GH1 gene polymorphism, which may confer lower levels of GH and IGF-I, appears to be associated with a decreased risk of colorectal cancer. (J Natl Cancer Inst 2002;94:454-60)