We conducted a study to determine the prevalence of central islands after refractive excimer laser surgery and the factors associated with their occurrence. A VISX Twenty/Twenty excimer laser was used to perform 157 photorefractive keratectomy (PRK) and 263 photorefractive astigmatic keratectomy (PARK) procedures. Corneal topography was assessed in 156 patients three months postoperatively using a topographic analysis system. Central islands were seen in 67% of patients; 26% had islands of 3.0 diopters (D) or more in height. Occurrence was associated with the laser suction nozzle being fixed during the procedure (P < .05). There was no demonstrable effect of central islands on visual acuity or refractive outcome, nor was there any relation to the amount of attempted correction or type of procedure (PARK or PRK). Central islands with a curvature difference of less than 3.0 D were seen more frequently than an absence of islands, suggesting these may be a "normal" postoperative topographic variation. The effect of nozzle position on island formation suggests that plume removal or corneal hydration may be important, as is corneal epithelial healing.