Schizophrenic illness is associated with high rates of osteoporosis, the etiology of which remains obscure, but which may be at least partly explained by the prolactin-raising properties of antipsychotic medication. Conventional antipsychotics all cause hyperprolactinemia, whereas a limited number of atypical antipsychotic drugs do not. To investigate this further, we designed a cross-sectional comparison study between groups taking either prolactin-raising or prolactin-sparing antipsychotic medication. Participants were required to be premenopausal women with a diagnosis of schizophrenia, and to have received exclusively either prolactin-raising (n = 26), or olanzapine (n = 12) antipsychotic medication. Half of the subjects in the prolactin-raising group were being treated with conventional (n = 13), and half with newer "atypical," antipsychotic drugs (n = 13). Subjects had lumbar spine and hip bone mineral density (BMD) evaluated by a dual-energy x-ray absorptiometer (DEXA) scan. A blood sample was taken to measure prolactin and sex hormone axis measures. The results demonstrated that the group taking prolactin-raising medication had higher rates of bone pathology, compared with the olanzapine group. High prolactin levels were related to measures of hypogonadism and low BMD values. Within the prolactin-raising group, those taking newer atypical compounds had higher levels of prolactin, lower levels of sex hormones, and lower BMD values than the group taking conventional antipsychotic medication. These findings suggest that the high rates of osteoporosis associated with schizophrenia may result from hypogonadism secondary to antipsychotic-induced hyperprolactinemia, and that the prolactin-raising profile of antipsychotic drugs should be considered when choosing an antipsychotic drug.