OBJECTIVE. The purpose of this study was to compare the efficacies of mammography, sonography, CT, and MR imaging in the detection of breast implant rupture and to analyze the imaging findings. SUBJECTS AND METHODS. Thirty-two women with 63 silicone breast implants participated in the study. All but one had signs and symptoms suggestive of rupture, and all had requested that their implants be removed before they were enrolled in this imaging study. All patients had film-screen mammography, sonography, CT, and MR imaging. Twenty-two ruptures were found at surgery; 21 were intracapsular and one was extracapsular. The relative efficacies of the imaging studies were determined, and the imaging findings were compared with the surgical results. RESULTS. Of the 32 women with 63 implants, mammographic sensitivity for detecting implant rupture was only 23% but the specificity was 98%. Sonography had a higher sensitivity (59%), but its specificity was significantly lower (79%). CT had a sensitivity of 82% and a specificity of 88%. MR was the only imaging technique that consistently provided evidence that enabled the evaluation of intracapsular and extracapsular ruptures. The sensitivity and specificity of MR imaging were 95% and 93%, respectively. CONCLUSION. Our results show that MR imaging is more sensitive and specific for the detection of breast implant rupture than is mammography, CT, or sonography.