We studied the effects of BRCA2 and CHEK2 variants on breast cancer risk in two case-control series from Poland and Belarus. The missense BRCA2 variant T1915M was associated with a significant reduction in breast cancer risk (OR = 0.62; 95% CI 0.49-0.79; P = 0.0007). Modest increases of breast cancer risk were observed for the four analysed CHEK2 variants (I157T, 1100delC, IVS2 + 1G > A and del5395) (OR = 2.2; 95% 1.7-2.8; P = 0.0001). The highest risk was observed among women who carried both a BRCA2 and a CHEK2 variant (OR = 5.7; 95% CI 1.7-19; P = 0.006). We observed a statistically significant interaction effect between CHEK2 mutations and the BRCA2 substitution (P = 0.03). These data suggest that the BRCA2 T1915M polymorphism alone might be associated with a reduced risk of breast cancer, but among CHEK2 mutation carriers, it may lead to an unexpectedly high risk.