The effects of estrogens are mediated primarily through estrogen receptor (ER) in breast tissue, and polymorphisms in the ER genes may alter the functions of these receptors. Polymorphisms in the ER-alpha gene have been reported to be associated with breast cancer risk. However, to our knowledge, no study has been published on the relation between ER-alpha gene polymorphisms and breast cancer survival.To determine whether three common polymorphisms in the ER-alpha gene, PvuII, XbaI, and GT dinucleotide repeats are associated with breast cancer survival, we evaluated data from a cohort of 1,069 breast cancer patients who participated in the Shanghai Breast Cancer Study between 1996 and 1998. The median follow-up time for this cohort of women was 5.2 years.No overall association was observed between ER gene polymorphisms and breast cancer survival. The genotype associations, however, were modified by ER status in breast cancer tissues. Comparing those with the PP genotype to the pp genotype of the PvuII polymorphism, the hazard ratios (HR) of dying were 3.30 [95% confidence interval (95% CI), 1.42-7.69] and 0.54 (95% CI, 0.24-1.23), respectively, for participants with ER-negative breast cancer and ER-positive breast cancer. Similarly, compared with those with no (GT)(23) alleles, carrying one or two (GT)(23) alleles of the GT repeat polymorphism was related to a HR of 1.48 (95% CI, 0.77-2.87) for ER-negative breast cancer and a HR of 0.25 (95% CI, 0.09-0.69) for ER-positive cancer. The effect of ER on breast cancer survival was also modified by genotypes of ER-alpha gene. Tests for multiplicative interaction were highly significant.These data suggest that the ER-alpha gene polymorphisms and ER status may have an interactive effect on breast cancer survival.