This study examined the effect of a 12-week non-instrumental resistance training program using body weight as a load (RT-BW) on body composition, fat distribution and metabolic profiles in elderly males and females. Healthy, non-diabetic, elderly volunteers (22 males and 30 females) aged 65-82 years were non-randomly divided into RT-BW (12 males and 20 females) and control (10 males and 10 females) groups. The RT-BW subjects were trained three times per week for 12 weeks according to a specified protocol involving a combination of upper and lower body weight and rubber tubing exercises. We evaluated body composition and fat distribution using anthropometry, dual energy X-ray absorptiometry and ultrasonography, and measured serum lipid levels and HbA(1c) at baseline and after 12 weeks of training. Changes over 12 weeks were significantly greater in the RT-BW group compared with the control group, with a decrease in waist circumference, pre-peritoneal (visceral) fat thickness and thigh fat thickness, and an increase in thigh muscle thickness. On the other hand, the changes in body weight, fat mass and fat free mass were no different between the groups. Further, there were significantly greater changes of metabolic profiles in the RT-BW group with an increase in HDL cholesterol and a decrease in triglyceride and HbA(1c). There was a significant between-group difference in diastolic blood pressure. Relatively short-term, non-instrumental resistance training using body weight as a load was effective in improving fat distribution and metabolic profiles in healthy elderly people without weight loss.