This study aimed to determine whether cosmetic breast implants impair the early detection of breast cancer, and adversely influence survival. This analysis derives from a cohort of 24,558 women who received bilateral cosmetic breast implants, and 15,893 women who underwent other plastic surgery procedures at the same practices in Ontario and Quebec, Canada, between 1974 and 1989. Incident cancers and vital status through 1997 were determined by record linkage to the Canadian Cancer Registry and Canadian Mortality Database. Analyses are based on a total of 182 and 202 incident cases of breast cancer identified among the implant and control groups, respectively. Contingency table analyses were performed to test for differences in the stage distribution of breast cancers between the 2 groups. Potential differences in survival were evaluated using the Kaplan-Meier estimates and Cox proportional hazards models. Women who received breast implants were more likely to have advanced stage breast carcinoma relative to the other plastic surgery patients (crude and adjusted ps </= 0.01). No statistically significant differences in distributions between the implant and control patients were found for age at diagnosis, tumor size, histological type, period of diagnosis or length of follow-up. The delayed diagnosis in augmented women did not appear to influence the overall prognosis. Breast cancer-specific survival was similar in both groups (hazard ratio = 1.06; 95% confidence interval = 0.65-1.74). In conclusion, this study suggests that breast implants delay the detection of breast cancer, but there was no statistically significant difference in survival between the breast implant and other plastic surgery groups.