Basic fibroblast growth factor (bFGF) has been shown to promote wound healing. The present trial evaluated the clinical efficacy of bFGF for diabetic ulcer, a type of refractory skin ulcer, and the dose-response relationship. This was designed as a randomized, double-blind, dose-ranging, placebo-controlled trial. A total of 150 patients with non-ischaemic diabetic ulcers measuring 900 mm2 or less were randomized into a placebo group (n = 51), a 0.001% bFGF group (n = 49) and a 0.01% bFGF group (n = 50), and 148 of these patients received treatment for 8 weeks or less. The efficacy evaluation was carried out on 139 patients who met the protocol in this trial. The primary outcome was the percentage of patients showing 75% or greater reductions in the area of ulcer. The area of ulcer decreased by 75% or more in 57.5% (27/47), 72.3% (34/47), and 82.2% (37/45) in the placebo, 0.001% bFGF and 0.01% bFGF groups, respectively, and differences were significant between the 0.01% bFGF and placebo groups (p = 0.025). The cure rate was 46.8% (22/47), 57.4% (27/47), and 66.7% (30/45) in the placebo, 0.001% bFGF and 0.01% bFGF groups, respectively. The findings obtained in this trial showed wound healing accelerating effects of bFGF on diabetic ulcers.