A relationship between circulating insulin and bone mass has been postulated. To test this hypothesis we measured fasting serum insulin and quantitated bone density in the forearm, spine and femoral neck by single and dual photon absorptiometry of 70 normal women. There was no statistically significant relationship between insulin and bone density at any site by correlation or regression analysis. Although the relationship between a single measurement of a hormone (at one point in time) and bone density (the result of lifelong factors) should be interpreted with caution, our findings do not support the hypothesis that there is an association between serum insulin and bone mass.