During the past 15 years, 194 penetrating keratoplasty specimens have been seen at the Eye Pathology Laboratory of the University of California Medical Center, San Francisco. Of those 194 specimens, 40 met with the following three criteria: (1) Each was obtained from an eye that had had at least one prior unsuccessful penetrating keratoplasty. (2) In each case, the diameter of the second keratoplasty was larger than the first, indicating that the specimen represented an excisional biopsy. (3) Each case was one for which clinical and histologic data were available. These 40 cases form the basis for this report on the correlation of clinical and histologic findings of unsuccessful penetrating keratoplasties. The 40 patients were all operated on at the University of California Hospital, San Francisco, and their cases represent the techniques of several surgeons, including the resident staff. During the 15-year period covered in this report, there have been