From 1986 to 1991, among 980 penetrating keratoplasties, all 22 patients were studied retrospectively who had undergone 28 refractive procedures because of high corneal astigmatism after penetrating keratoplasty. In 8 procedures, only relaxing incisions were made, and in 20 procedures, additional compression sutures were placed at 90 degrees to the former. In 7 cases the relaxing incisions were located on the patient's peripheral cornea and in 21 cases, inside the graft margin. The absolute preoperative corneal astigmatism was 7.25-20.0 D (mean, 11.5 +/- 3.1 D). Immediately after surgery it was 0-14.0 D (mean, 7.0 +/- 3.6 D). The vector-corrected astigmatism immediately after surgery was 1.0-28.9 D (mean, 13.2 +/- 7.8 D). After a mean follow-up of 20.6 months, the mean residual astigmatism was 1.0 +/- 11.25 D (mean, 5.4-2.5 D). The mean postoperative vector-corrected astigmatism was 0-19.3 D (mean, 9.1 +/- 4.6 D). The best-corrected preoperative visual acuity was 0.12-0.9 (mean, 0.4), the best postoperative visual acuity was 0.3-1.2 (mean, 0.6). The approach described may reduce disturbing postkeratoplasty astigmatism--although the predictability remains unsatisfactory.