We propose a hypothesis that differences in base excision repair capacity modulate the effect of dietary antioxidant intake on prostate cancer risk. As a preliminary test of this hypothesis, we conducted a pilot case-control study to evaluate prostate cancer risk in men with polymorphisms in the XRCC1 gene, a key player in base excision repair, across different strata of antioxidant intake. Seventy-seven prostate cancer patients and 183 community controls, for whom we have detailed dietary information, were frequency matched on age and race. We found a somewhat lower prostate cancer risk for men with one or two copies of the variant alleles at the XRCC1 codons 194 and 399 than for those who were homozygous for the common allele [codon 194: odds ratio (OR) = 0.8; 95% confidence interval (CI), 0.4-1.8 and codon 399: OR = 0.8; 95% CI, 0.5-1.3]. The variant at codon 280 was associated with a slightly increased prostate cancer risk (OR = 1.5; 95% CI, 0.7-3.6). Only the codon 399 polymorphism occurred frequently enough to investigate its joint effect with antioxidant intake. Prostate cancer risk was highest among men who were homozygous for the common allele at codon 399 and had low dietary intake of vitamin E (OR = 2.4; 95% CI, 1.0-5.6) or lycopene (OR = 2.0; 95% CI, 0.8-4.9), whereas low intake of these antioxidants in men without this genotype hardly increased prostate cancer risk. The polymorphism did not modulate risk associated with low intake of vitamin C, A, or beta-carotene. The data give some support for our hypothesis but should be regarded as preliminary, because it is limited by small sample size. We discuss what kind of data and what kind of studies are needed for future evaluation of this hypothesis.