AIM: To analyze the concentration-dependent effects of autologous serum(AS) and fetal bovine serum(FBS) on human corneal epithelial cell(HCEC) viability, migration and proliferation.METHODS: AS was prepared from 13 patients with nonhealing epithelial defects Dulbecco’s modified eagle medium/Ham’s F12(DMEM/F12) with 5% FBS, 0.5% dimethyl sulphoxide(DMSO), 10 ng/m L human epidermal growth factor, 1% insulin-transferrin-selenium, then were incubated in serum media: DMEM/F12 supplemented by 5%, 10%, 15% or 30% AS or FBS. HCEC viability was analyzed using cell proliferation kit XTT, migration using a wound healing assay, proliferation by the cell proliferation enzyme-linked immunosorbent assay(ELISA) Brd U kit. Statistical analysis was performed using the generalized linear model, the values at 30% AS or 30% FBS were used as the baselines.RESULTS: HCEC viability was the highest at 30% AS or 15% FBS and the lowest at 10% AS or 30% FBS application. HCEC migration was the quickest through 30% AS or 30% FBS and the slowest through 5% AS or 5% FBS concentrations. Proliferation was the most increased through 15% AS or 5% FBS and the least increased through 30% AS or 30% FBS concentrations. HCEC viability at 10% and 15% AS was significantly worse(P=0.001, P=0.023) compared to baseline and significantly better at 15% FBS(P=0.003) concentrations. HCEC migration was significantly worse(P≤0.007) and HCEC proliferation significantly better(P<0.001) in all concentration groups compared to baseline.CONCLUSION: For the best viability of HCEC 30% AS or 15% FBS, for HCEC migration 30% AS or 30% FBS, for proliferation 15% AS or 5% FBS should be used. Therefore, we suggest the use of 30% AS in clinical practice.