The results of excimer laser photorefractive keratectomy (PRK) in 16 blind and 120 sighted eyes (136 patients) are presented. Follow up for the blind eyes was 22 months and for the sighted eyes 8 to 18 months (mean, 1 year). A Summit Technology UV200 excimer laser with fluence fixed at 180 mJ/cm2, beam diameter 4 mm, a frequency of 10 Hz, and ablation rate of 0.22 mu per pulse was used. The range of preoperative myopia was - 1.50 diopters to - 17.50 D (spherical equivalent). Ninety percent of patients undergoing a -2.00 D correction and 75% of patients undergoing a -3.00-diopter correction were within +/- 1.00 D of intended refraction at 6 months. This figure fell to 40% and 20% for the groups undergoing -6.00 D and -7.00 D corrections, respectively. While the trend is toward undercorrection, around 70% of patients in all groups have benefited from a reduction of their myopia by at least half that intended. A variable degree of anterior stromal "haze" was detected in 110 patients (92%) despite a rigid high dose topical corticosteroid regime over 3 months. Best spectacle corrected visual acuity, however, was greater than or equal to preoperative levels in 111 patients (93%). Six patients experienced a reduction of one line of Snellen acuity at 6 months and three of the high myopes lost two lines. No patients lost more than two lines of Snellen acuity. Ninety-four patients (78%) noticed a "halo" around lights at night, more marked in the early postoperative period, which we attribute to the 4-millimeter diameter ablation zone. With a large pupil, both central and paracentral cornea can contribute simultaneously to the retinal image. We have demonstrated considerable individual variation in response to this surgery, which is more marked in high myopia. Although follow up at the present time is limited, to date PRK appears to be a safe procedure that holds considerable promise for refractive surgery in the future.