OBJECTIVE: To assess, in obese type 2 diabetics (T2D), the impact of a home-based effort training program and the barriers to physical activity (PA) practice. METHOD: Twenty-three obese T2D patients (52.7 +/- 8.2 years, BMI = 38.5 +/- 7.6 kg/m(2)) were randomized to either a control group (CG), or an intervention group (IG) performing home-based cyclergometer training during 3 months, 30 min/day, with a monthly-supervised session. The initial and final measurements included: maximal graded effort test on cyclergometer, 6-minute walk test (6MWT) and 200-meter fast walk test (200mFWT), quadriceps maximal isometric strength, blood tests and quality of life assessment (SF- 36). A long-term assessment of the amount of physical activity (PA) and the barriers to PA practice was conducted using a questionnaire by phone call. RESULTS: Patients in the CG significantly improved the maximal power developed at the peak of the cyclergometer effort test (P < 0.05) as well as the quadriceps strength (P < 0.01). There were no significant changes in the other physical and biological parameters, neither in quality of life. At a mean distance of 17 +/- 6.4 months, the PA score remained low in the two groups. The main barriers to PA practice identified in both groups were the perception of a low exercise capacity and a poor tolerance to effort, lack of motivation, and the existence of pain associated to PA. CONCLUSION: This home-based intervention had a positive impact on biometrics and physical ability in the short term in obese T2D patients, but limited effects in the long term. The questionnaires completed at a distance suggest considering educational strategies to increase the motivation and compliance of these patients.