Patients with diabetes have been shown to suffer from increased fall risk. However, authors disagree as to whether only diabetic patients with neuropathy, or also those without neuropathy, present gait alterations. Existing studies evaluate gait indoors, i.e. in specialized gait laboratories. This study evaluates gait parameters in diabetic patients under various real life conditions and compares them to those recorded for healthy controls. METHODS: We conducted a clinical observation study. Forty-five subjects' gait was assessed on three different surfaces (tar, grass and stones) with a Physilog system (BioAGM, CH), consisting of accelerometers and gyroscopes. Temporal and spatial gait parameters as well as stride-to-stride variability of 30 patients with type 2 diabetes, 15 with and 15 without neuropathy were compared to 15 healthy controls. The three groups were comparable for age, height and body weight (p>0.05). RESULTS: Diabetic patients' gait parameters differed significantly from those of healthy controls. Post hoc analysis revealed a significant difference between healthy individuals and patients with neuropathy, and between healthy individuals and patients without neuropathy. No difference was observed between patients with and without neuropathy. The highest surface effect was found in patients with diabetic neuropathy, followed by patients without neuropathy and healthy controls. CONCLUSIONS: Walking in real life conditions revealed gait difficulties in patients with type 2 diabetes before neuropathy was clinically detectable. Clinicians should be aware that diabetic individuals' gait capacity decreases and fall risk increases at an early stage of the disease.