The aim of this study was to investigate the relationship between the level of stromal surface irregularity after photorefractive keratectomy (PRK) and myofibroblast generation along with the development of corneal haze. Variable levels of stromal surface irregularity were generated in rabbit corneas by positioning a fine mesh screen in the path of excimer laser during ablation for a variable percentage of the terminal pulses of the treatment for myopia that does not otherwise generate significant opacity. Ninety-six rabbits were divided into eight groups: [see table in text]. Slit lamp analysis and haze grading were performed in all groups. Rabbits were sacrificed at 4 hr or 4 weeks after surgery and histochemical analysis was performed on corneas for apoptosis (TUNEL assay), myofibroblast marker alpha-smooth muscle actin (SMA), and integrin alpha4 to delineate the epithelial basement membrane. Slit-lamp grading revealed severe haze formation in corneas in groups IV and VI, with significantly less haze in groups II, III, and VII and insignificant haze compared with the unwounded control in groups I and V. Analysis of SMA staining at 4 weeks after surgery, the approximate peak of haze formation in rabbits, revealed low myofibroblast formation in group I (1.2+/-0.2 cells/400x field) and group V (1.8+/-0.4), with significantly more in groups II (3.5+/-1.8), III (6.8+/-1.6), VII (7.9+/-3.8), IV (12.4+/-4.2) and VI (14.6+/-5.1). The screened groups were significantly different from each other (p < 0.05), with myofibroblast generation increasing with higher surface irregularity in the -4.5 diopter PRK groups. The -9.0 diopter PRK group VI had significantly more myofibroblast generation than the -9.0 diopter PRK with PTK-smoothing group VII (p < 0.01). Areas of basement membrane disruption were demonstrated by staining corneas for integrin alpha4 and were prominent in corneas with grade I or higher haze. SMA-positive myofibroblasts tended to be present sub-adjacent to basement membrane defects. Late apoptosis was detected at 1 month after surgery within clusters of myofibroblasts in the sub-epithelial stroma. In conclusion, these results demonstrated a relationship between the level of corneal haze formation after PRK and the level of stromal surface irregularity. PTK-smoothing with methylcellulose was an effective method to reduce stromal surface irregularity and decreased both haze and associated myofibroblast density. We hypothesize that stromal surface irregularity after PRK for high myopia results in defective basement membrane regeneration and increased epithelium-derived TGFbeta signalling to the stroma that increases myofibroblast generation. Late apoptosis appears to have a role in the disappearance of myofibroblasts and haze over time.