The study aim was to investigate the health effects of a pedometer-based behavioural modification program in type 2 diabetes patients and to examine the relationship between changes in steps/day (baseline-post and baseline-follow up) and health outcomes. Ninety-two type 2 diabetes patients (69% male, mean age: 62 ± 9 years and mean BMI: 30.0 ± 2.5 kg/m(2)) were recruited and randomly assigned to an intervention or control group. The intervention consisted of one face-to-face session, pedometer use and seven telephone calls. Selection criteria included 35-75 years, 25-35 kg/m(2) and ≤12% HbA1c (108 mmol/mol). Outcome measures were assessed at baseline, post and follow up, and included systolic blood pressure, waist circumference, body mass index, glucose control (HbA1c and fasting glucose), triglycerides, total, HDL and LDL cholesterol and steps/day. The results showed no significant short- or intermediate-term differences in health outcomes between the control and intervention group. However, a threshold was identified, as HbA1c improved significantly in those who increased ≥4000 steps/day between baseline- and post-measurements (n = 18). This threshold was not applicable to any other health outcome. Hence, although the intervention successfully increased steps/day, no direct effect on health outcomes was identified. However, an increase of ≥4000 steps/day seemed a threshold to have a positive impact on HbA1c.