AIM: To identify clinical factors associated with gait alterations in patients with Type 2 diabetes. METHODS: A sample of 76 diabetic patients underwent clinical examination and an outdoor gait assessment on tarred and cobblestoned terrains. We calculated respective differences in gait speed (performance measure) and gait variability (fall risk index) on changing terrains. Associations with clinical factors were investigated using correlation coefficients and linear regression analysis. RESULTS: The mean walking speed on the tarred pathway was 4.5 +/- 0.6 km/h and 3.9 +/- 0.8 km/h on the cobblestone pathway (P < 0.001). The CVGCT increased from 2.6 +/- 0.9% on the tarred pathway to 5.1 +/- 2.8% on the cobblestone pathway (P < 0.001). Regression analysis showed that 36% of the decrease in gait speed was explained proportionally by the mean of maximal isometric lower limb strength (22.2%; P ≤ 0.01), fear of falls (7.4%; P ≤ 0.01) and participants' perceived vibration threshold (6.4%; P ≤ 0.01). Moreover, mean maximal isometric strength explained 11.8% (P ≤ 0.01) of the increase of the coefficient of variation of the gait cycle time when participants changed from tarred terrain to cobblestones. CONCLUSION: This study indicated that both physiological (strength and proprioception) and cognitive-behavioural factors (fear of falls) should be considered when treating diabetic patients with gait alterations. Therapists should apply these findings when developing specific fall prevention and treatment programmes.