This epidemiologic study was undertaken to determine whether working overtime is associated with anthropometric indices and serum lipids, risk factors for obesity, in white-collar workers. Non-management white-collar male workers were eligible. Body weight and height, waist circumference, skinfold thickness, serum total cholesterol and triglycerides were measured. Weight, height and waist circumference data obtained 3 years previously were also used. Lifestyle information was obtained by means of a self-administered questionnaire. Overtime hours correlated significantly with the 3-year change in body mass index (r = 0.206, p < 0.0017) and waist circumference (r = 0.218, p = 0.0091), but not with either the most recent anthropometric indices or serum lipids. Overtime hours were also intercorrelated (r = 0.436, p < 0.0001) with dinner time. The present study suggested that working overtime is associated with the increases in BMI and waist circumference over a 3-year period although the associations were weak. Additionally, eating habits of those with working overtime might reflect an intervening effect on the anthropometric changes.