Corticosteroids are currently administered after photorefractive keratectomy (PRK) to reduce corneal haze and myopic regression. However, the beneficial effects of corticosteroids are controversial, and they are associated with many side effects. In this study we compared the in vitro antiproliferative effects of nonsteroidal antiinflammatory drugs (NSAIDs), diclofenac and flurbiprofen, with those of dexamethasone on human keratocytes. Human keratocytes were incubated with various concentrations of diclofenac, flurbiprofen, and dexamethasone. The control samples were incubated under the same conditions except for the absence of drugs. Proliferation of the keratocytes was measured by [3H]thymidine uptake into DNA on days 1, 2, and 4. Diclofenac was the most potent agent, inducing dose-dependent inhibition at concentrations ≥101 mM on days 1 and 2, and ≥10-1 mM on day 4. Flurbiprofen followed closely, inhibiting keratocytes at and above 1 mM on day 1, 10-1 mM on day 2, and 10-4 mM on day 4. Dexamethasone was the least effective, exhibiting inhibition at and above 25 mM on day 1, 5 mM on day 2, and 1 mM on day 4 (Wilcoxon rank sum test, p ≤0.05). The ID5Oδ reflect the same trend (day 4: diclofenac=0.03 mM, flurbiprofen=0.2 mM, dexamethasone=3.2 mM). In addition, diclofenac and dexamethasone showed time-dependent antiproliferative effects. These results indicate that NSAIDs are more potent than corticosteroids in inhibiting proliferation of human keratocytes in vitro, and suggest a potential use of NSAIDs in modulating corneal wound healing after PRK