A clinically, immunohistochemically and ultrastructurally characterized series of 192 pituitary adenomas was analyzed for DNA content by flow cytometry. Results were assessed not only relative to tumor immunotype, size, and invasiveness, but also with frequency of recurrence. Case selection was non-random; males predominated (1.8:1) and the ratio of macro-to-microadenomas was 4.2:1. Female patients were slightly younger and, in all adenoma categories, less often had invasive tumors: PRL (15%/30%), ACTH (17%/44%), LH/FSH (8%/27%) and null cell adenomas (0%/27%). With the exception of prolactin cell adenomas, similar proportions of macroadenomas and invasive tumors in all tumor subtypes were diploid and non-diploid. Prolactin adenomas differed in that tumors of males showed a high rate of non-diploidy (65%); such tumors were predominantly macroadenomas, but only 28% were invasive. Among GH-containing tumors 78% were macroadenomas, 40% were nondiploid, and the frequency of invasive macroadenomas was higher (49%) than in PRL tumors (21%). ACTH adenomas were mainly microadenomas (81%), their rate invasion (29%) and of non-diploidy being low (14%). Among "non-functioning" (LH/FSH, null cell adenomas), LH/FSH-producing tumors were all macroadenomas, but with low rates of invasion (23%) and non-diploidy (9%). Null cell adenomas, nearly all macroadenomas, had similar low invasion rate (21%), but were more often non-diploid (39%). In all adenoma subgroups S-phase fractions were higher in non-diploid adenomas by an overall ratio of 2.1:1. Prolactin adenomas showed the highest (15.2%) and LH/FSH adenomas the lowest (5.6%) mean S-phase fraction. When compared to long-term follow-up, neither this parameter nor ploidy correlated with tumor size or invasiveness. Lastly, long-term follow-up showed ploidy to be an unreliable predictor of tumor persistence or recurrence.