Aims/hypothesis Walking is recommended as an adjunct therapy to diet and medication in diabetic patients, with the aim of improving physical fitness, glycaemic control and body weight reduction. Therefore we evaluated walking activity on the basis of capacity, performance and potential risk of plantar injury in the diabetic population before it can be prescribed safely. Subjects, materials and methods Twenty-three subjects with diabetic neuropathy (DMPN) were compared with 23 patients with current diabetic foot ulcers, 16 patients with partial foot amputations and 22 patients with trans-tibial amputations. The capacity for walking was measured using a total heart beat index (THBI). Gait velocity and average daily strides were measured to assess the performance of walking, and its impact on weight-bearing was studied using maximum peak pressure. Results THBI increased (p<0.01) and gait velocity and daily stride count fell (p<0.001 for both) with progression of foot complications. The maximum peak pressures over the affected foot of patients with diabetic foot ulcers (p<0.05) and partial foot amputations (p<0.01) were higher than in the group with DMPN. On the contralateral side, the diabetic foot ulcer group showed higher maximum peak pressure over the total foot (p<0.05), and patients with partial foot amputations (p<0.01) and trans-tibial amputations (p<0.05) showed higher maximum peak pressure over the heel. Conclusions/interpretation Walking capacity and performance decrease with progression of foot complications. Although walking is recommended to improve fitness, it cannot be prescribed in isolation, considering the increased risk of plantar injury. For essential walking we therefore recommend the use of protective footwear. Walking exercise should be supplemented by partial or non-weightbearing exercises to improve physical fitness in diabetic populations.