A STUDY WAS MADE AT MASSACHUSETTS CORRECTIONAL INSTITUTION OF CONSECUTIVE PRISONERS (89 WOMEN, 11 MEN) REFERRED FOR PSYCHIATRIC EVALUATION. THE STUDY WAS INITIATED WHEN IT BECAME EVIDENT THAT A NUMBER OF INMATES WITH ACUTE MANIA HAD NOT BEEN DIAGNOSED PREVIOUSLY AS HAVING SUCH DISORDERS. A COMPARATIVELY HIGH INCIDENCE (10 PERCENT) OF PRIMARY AFFECTIVE DISORDER WAS FOUND IN THIS STUDY, WHICH HYPOTHESIZED THAT MANIC DEPRESSIVE STATES MAY BE UNDERDIAGNOSED IN THE UNITED STATES. SEVEN ADDITIONAL PATIENTS HAD EQUIVOCAL, DEFERRED, OR SCHIZOAFFECTIVE DIAGNOSES. SEVEN OF THE PATIENTS WITH PRIMARY AFFECTIVE DISORDER HAD BIPOLAR MOOD SWINGS, TWO WERE UNIPOLAR OR WITHOUT DEFINITE EVIDENCE OF MANIA, AND ONE WAS UNIPOLAR WITH POSSIBLE BIPOLARITY. SIGNIFICANTLY, IN ONLY TWO CASES WAS THE AFFECTIVE DIAGNOSIS KNOWN OR AVAILABLE AT THE TIME OF COMMITMENT TO PRISON; IN ONE CASE DIAGNOSIS WAS MADE AFTER COMMITMENT WHEN A HISTORY WAS GIVEN. IN THE REMAINING SEVEN CASES, EITHER NO PREVIOUS DIAGNOSIS WAS AVAILABLE OR THE DIAGNOSIS MERITED REVISION. SEVERAL FACTORS MAY CONTRIBUTE TO UNDERDIAGNOSIS OF PRIMARY AFFECTIVE DISORDER. AMONG THESE ARE MASKED DIAGNOSIS (EXCLUSION OF PATIENTS FROM DIAGNOSIS BECAUSE OF PAST ACTIONS OR THEIR IDENTIFICATION AS PRISONERS), MISLABELING OF THE DISORDER LEADING TO INCORRECT DIAGNOSIS, AND INSTITUTIONAL COMPARTMENTALISM AND AND INADEQUATE LIAISON BETWEEN PRISONS AND PSYCHIATRIC WORKERS. THE REPORT INCLUDES CASE HISTORIES OF THE 10 MANIC DEPRESSIVE PATIENTS, TABULAR DATA, AND A LIST OF REFERENCES. (MJW)