Postoperative changes in corneal transparency is a major concern after PRK. Data were obtained from 69 human eyes treated with excimer laser photorefractive keratectomy in order to evaluate a relationship between objective measurements of corneal transparency and visual performance. A CCD-camera device was employed and by using polarizing filters the system could discriminate between reflected and scattered light. We observed two groups of postoperative behaviour in corneal transparency. Each group showed an increase in both scattered and reflected light signals with a maximum at around two months postoperatively. In the majority group (70%) this was followed by a subsequent decline of the scattered light signal, whereas the combined signal generated by reflected and scattered light showed a biphasic curve with a second peak at around 4 to 5 months postoperatively. In the minority patient group (30%) the timebase variations in pattern of both signals were indistinguishable throughout the period of observation. In all patients we observed a good correlation between the signal generated by scattered light alone and the reduction in the 5% contrast visual acuity performance, whereas correlation was poor when the combined signal of reflected and scattered light was considered. These disturbances in low contrast visual performance were only significant during the first three to four months postoperatively and thereafter most patients returned to their preoperative value. Eighteen percent of our patients discontinued topical steroids postoperatively. No differences in either corneal transparency or final refraction was observed. We consider that PRK is an effective form of refractive surgery and the marginal loss of corneal transparency should only be a problem during the first three or four months postoperatively.