A multicenter, double-blind, placebo-controlled trial was carried out to evaluate the intra-lesional infiltration of recombinant epidermal growth factor (EGF) in Wagner's grade 3 or 4 diabetic foot ulcers (DFUs). Subjects (149) were randomised to receive EGF (75 or 25 microg) or placebo, three times per week for 8 weeks and standard good wound care. The main endpoint was granulation tissue covering > or = 50% of the ulcer at 2 weeks. It was achieved by 19/48 controls versus 44/53 in the 75 microg group [odds ratio (OR): 7.5; 95% confidence interval (CI): 2.9-18.9] and 34/48 in the 25 microg group (OR: 3.7; 1.6-8.7). Secondary outcome variables such as end-of-treatment complete granulation response (28/48 controls, 46/53 with 75 microg and 34/48 with 25 microg EGF), time-to-complete response (controls: 5 weeks; both EGF dose groups: 3 weeks), and wound closure after follow-up (25/48 controls, 40/53 with 75 microg and 25/48 with 25 microg EGF) were also treatment dependent. Multivariate analyses yielded that they were significantly enhanced by 75 microg EGF treatment and neuropathic versus ischemic ulcers. Most adverse events were mild and no drug-related severe adverse reactions were reported. It was concluded that recombinant human EGF (rhEGF) local injections offer a favourable risk-benefit balance in patients with advanced DFU.