to study the role of clinical, demographic and laboratory factors on the levels of sex steroids, we studied 28 men with type 2 diabetes (dm2) and 27 non-diabetic controls. body mass index (bmi), waist size and waist-hip ratio (whr) were measured in all, as well as, glucose, hemoglobin a1c, insulin, peptide c, estradiol, testosterone, and shbg. the free testosterone index (fti) was calculated by dividing total testosterone (tt) by shbg. the homa-r index was used as a measure of insulin resistance. we found that diabetics had a higher homa-r than non-diabetics (± vs. ±; p= ), a tendency toward lower levels of tt (3400 [1829-7000] vs. 4267 [2097-7074]pg/ml; p= ) and lower fti ( [-] vs. [-]; p= ). linear regression analysis showed that age was the only variable that affected fti (r= , r2= ; p= ) and that homa-r (r= , r2= -; p= ) and age (r= , r2= ; p= ) influenced shbg. we conclude that diabetes is partially responsible for lower levels of total and free testosterone in men matched for age and degree of body fat, but age was the most important determinant of lower level of androgens in men with and without diabetes.