In Reply—
The comments by Thompson and Weissman seem fair enough. Real data are difficult to come by and data concerning the possible etiologic effects of neuroleptic drugs on human breast cancer are scarce indeed. Although recognizing the importance of statistical tests, it seemed to me to be clinically relevant to report that, of some 3,278 women who were treated for schizophrenia in the University of Texas Medical Branch (Galveston) hospitals over a 15-year period, only six returned during the last decade of that period for diagnosis and/or treatment of breast cancer. The ratio of prior diagnosis of schizophrenia to all prior psychiatric diagnoses among women who were treated for breast cancer was almost identical to the ratio of schizophrenia to all psychiatric diagnoses in the general psychiatric population of the hospital during the period studied. Thus, patients who were previously treated for schizophrenia were not overly represented relative to other