OBJECTIVE In the present study, we used the diabetic mouse as a model of delayed wound healing to investigate the effects of human cord blood mesenchymal stem cells (CB-MSC) on wound healing. METHODS A delayed wound healing model was used by db/db mice. Study models were divided by an injection of human CB-MSC with phosphate buffered solution (PBS) by a different method. One was a locally topical injection, the other was a systemic injection via the end tail vein. Both models were treated with 2.0 × 10(6) CB-MSC after an 8-mm full thickness defect was made by a skin punch biopsy on the back. We evaluated the wound size, transforming growth factor (TGF)-β, and vascular endothelial growth factor histologic evaluation, and vessel counts. Engraft of CB-MSC was detected by an antihuman antibody. RESULTS Wound healing was accelerated in the experimental group in the topical injection model with statistical significance on the 6th, 9th, and 12th day (P < 0.05). In the systemic injection model, wound healing was completed from the 9th day, but there was no statistical significance. TGF-β increased in the first week and decreased in the third week in the experimental groups of both models. But there were opposite results in the control groups of both models. The statistical differences were found in first and third week in topical injection and in the third week in systemic injection (P < 0.05). Vascular endothelial growth factor increased in all groups and in all models as the wound healing. But statistical significance did not show between all experimental and control groups. Anti-human antibody immunochemical staining was positive in the wound. CONCLUSIONS We concluded that CB-MSC had a positive effect on wound healing. Statistically significant results were noted in the topical injection model. We also reported good effects on the systemic injection model, although we did not find any statistical significance. CB-MSC may influence wound healing by TGF-β.