There have been several anecdotal reports that silicone breast implants are associated with an increased incidence of autoimmune disease. Based upon these data as well as the theoretical potential of silicon and silicone immune interactions, we hypothesized that an immune response to a silicone breast implant would include host reactivity against components of the microenvironment within the implant milieu. To test this hypothesis, we obtained detailed histories and performed examinations of 57 consecutive, self-referred patients concerned about their breast implants. Eleven of these women were excluded for various reasons including previous exposure to bovine collagen. The remaining 46 women, as well as 45 normal women of approximately the same age and living in the same geographic region, were tested using a sensitive ELISA for the presence of autoantibodies to human native type I collagen, denatured type I collagen, native type II collagen and denatured type II collagen. Known positive and negative sera were included in all assays and the ELISA was performed and interpreted blindly. Positive sera were defined as an ELISA value of three standard deviations above the mean of the normal controls. Using these stringent criteria, there was a statistically significant incidence of antibodies to collagen in women with silicone breast implants. In fact, 35% of women with silicone breast implants had such antibodies; this is higher than we have observed in any other autoimmune disease and is similar to that of chronic erosive rheumatoid arthritis. We believe that silicone breast implants, in genetically susceptible hosts, may pose a significant risk for immunopathology.