A weight reduction program to improve cardiovascular risk factors was implemented in obese subjects. The program consisted of exercise training corresponding to the anaerobic threshold (AT) and a mild hypocaloric diet for 12 weeks. In this program, we evaluated the effects of a combination of exercise training and a diet on cardiovascular risk factors such as obesity, dyslipidemia, and poor exercise performance in obese subjects. In addition, we also evaluated the independent effects of exercise training and dietary modification. For this purpose, we adopted a relative training time and a diet score. A relative training time was calculated as the number of times that the subject performed exercises divided by all of the training sessions scheduled, and the diet score was calculated from information which each subject provided on a self-assessment questionnaire. Twenty three obese subjects (Age: 24-54 years old, 19 men and 4 women, body mass index (BMI) > 26 kg/m2) participated in this study. After the 12-week intervention, the mean reductions in body weight, body mass index and body fat were 4.7 kg, 1.7 kg/m2 and 2.9%, respectively (P < 0.0001). The % change in body weight was significantly associated with the diet score and with the relative training time. The mean reductions in total cholesterol, triglyceride and low density lipoprotein cholesterol were 21 mg/dl (P < 0.002), 34 mg/dl (P < 0.01) and 15.9 mg/dl (P < 0.01), respectively, and the % change in triglyceride was significantly associated with the diet score (P = 0.0056) and tended to correlate with the relative training time (P = 0.0596). Oxygen uptake at AT and at peak exercise were increased from 14.1 +/- 1.6 to 16.0 +/- 3.1 ml/min/kg (P < 0.005) and from 26.3 +/- 4.8 to 28.4 +/- 4.9 ml/min/kg (P < 0.002), respectively. A combination of aerobic exercise and a mild hypocaloric diet significantly contributed not only to weight loss but also to the improvement of dyslipidemia and exercise performance, but either hypocaloric diet or mild exercise independently did less. The diet score and the relative training time were useful for evaluating separately dietary modification and the quantity of exercise.