To understand better the pathogenesis of pituitary chromophobe adenoma, we undertook a retrospective study of host factors, sexual function, associated diseases, and natural history. Data were obtained from the records of 464 patients in whom the disease was demonstrated by pathologic, radiographic, laboratory, and clinical means. The patients were drawn from the Presbyterian Hospital and the Neurological Institute of New York City; the study period comprised the years 1930–1964. The definitely positive findings are: peak age of onset 31–60 yr, a startlingly low (3%) prevalence of left-hand preference, a significantly earlier menopause (by 13 yr on average) than controls, a low frequency of marriage and a small number of offspring, the expected association with acromegaly (3.8%) and Cushing's syndrome (1.1%), an unexpectedly high frequency of diabetes mellitus (8.6%), perhaps in part responsible for the prevalence (7.5%) of coronary artery disease, a low rate of occurrence of fibromyomata uteri; the commonest chief complaint being visual symptoms in about two thirds of patients that sooner or later appear in nearly three quarters of the patients; and the much earlier onset (by years) of endocrinel than of neurologic dysfunction—particularly obvious in the form of early menopause. Suggestively positive findings are: a predilection of Jews and Old Americans for the disease, a somewhat low frequency of Rh negative individuals, a slightly early menarche, perhaps a high frequency of duodenal ulcer, and the “conventional” order of appearance of target gland failure—gonadal followed later by thyroidal and adrenocortical deficiency. This “classic” order of events should be viewed with caution because in this series as in earlier reports samples are small and data incomplete. The negative findings are: equal incidence in the two sexes (contrary to a few previous reports), absence of a distinctive physique, and normal frequency of ABO blood group distribution. Statistical analysis of the data demonstrates that large groups of patients are essential for accurate conclusions. Equally necessary are comparable control groups, with respect to era when studied, age, and racial mix. With regard to pathogenesis, the most promising leads to emerge from this study are (1) a careful genetic investigation is warranted to assess the frequent association with diabetes mellitus; and (2) the long period between earliest symptoms, which are usually endocrinel, and the neurologic symptoms that appear years later, suggests the need for a prospective laboratory study of patients suspected of harboring a chromophobe adenoma. Only in this way can one assess the importance of anterior pituitary target gland failure as a primary, not a secondary, event in the pathogenesis of chromophobe adenoma.