Recent studies suggest that cardiovascular disease is associated with abdominal distribution of adipose tissue rather than obesity in terms of total body fat. A number of other variables, known to be associated with obesity, were therefore examined in a cohort of randomly selected middle-aged men in relation to abdominal distribution of adipose tissue, measured as the ratio of the circumferences of the waist and hips (WHR), as well as to degree of obesity, measured as body mass index (BMI). These variables included anthropometric variables, cardiovascular risk factors as well as socioeconomic factors and physical health. Increased WHR, independent of BMI, was negatively associated with height, and hip circumference. Positive associations were found with blood pressure, cholesterol, triglycerides, fibrinogen and smoking. In addition positive associations were found with low social class and social group, illness in terms of sick leave, frequent use of health facilities such as X-rays, as well as diseases such as peptic ulcer. In sharp contrast to this, BMI, independent of WHR, was not associated with physical health variables or social class. Generalized obesity seemed to be associated with good health in the variables measured. There were positive associations to various anthropometric variables, including lean body mass. High BMI was also associated with elevated blood pressure and triglycerides. Several of the indicators of poor health traditionally associated with obesity thus do not seem to be characteristic for obesity in middle-aged men selected at random from the population but rather for an abdominal fat distribution, independent of obesity.