BACKGROUND Silicone augmentation mammoplasty has been postulated as a cause of environmentally-induced scleroderma. While representing a small proportion of all alleged causes of scleroderma, the issue has huge social, ethical and medicolegal ramifications. The hypothesis, however, has been recently questioned in results of comparative studies. We have previously reported no association between augmentation mammoplasty and scleroderma. However, all information was self-reported including augmentation mammoplasty status, and the prosthesis type was not identified. In addition, data were not available from untraceable cases. The current study addresses these issues. OBJECTIVE To validate self-reported augmentation mammoplasty status, re-analyse rates of exposure to silicone gel breast prostheses in 556 scleroderma patients and 289 general practice controls and evaluate whether silicone gel breast prostheses are causally linked to scleroderma. METHODS Study design-population-based case-control study; Cases-scleroderma patients resident in Sydney for at least six consecutive months between 1974-1988; Controls-patients from 29 randomly selected Sydney general practices, age- and gender-group-matched with cases. Validation of augmentation mammoplasty exposure was ascertained from the general medical practitioner of each interviewed case and control, or from the medical records of each deceased or untraceable case. Validation of the date of surgery and prosthesis type was from the relevant plastic surgeon. For each augmentation mammoplasty-positive case, validation of both the date and nature of the scleroderma onset was from the patient's medical records. Controls were given a "control date' for disease onset to adjust for duration of potential exposure. RESULTS Validation of augmentation mammoplasty status was possible in 252 (87.2%) living controls, and 532 (95.7%) cases, of whom 287 were living. Self-reported augmentation mammoplasty status was highly reliable in living non-senile cases (kappa = 1), and living validated controls (kappa = 0.86). No association was identified between silicone gel augmentation mammoplasty and scleroderma with unadjusted odds ratios of 1.33 (95% CI: 0.26-6.71), and 1.00 (95% CI: 0.16-6.16) following adjustment for potential confounders of age, socioeconomic status and ethnicity. CONCLUSIONS This validates the self-reported augmentation mammoplasty status previously reported and does not support the hypothesis that silicone gel augmentation mammoplasty is an environmental inducer of scleroderma in females.