BACKGROUND Although several studies have documented the existence of psychopathology in the morbidly obese, there is disagreement as to its extent and nature. The disagreement has been difficult to resolve because earlier studies have tended to use small, regional samples, and diverse, unstandardized approaches to measuring psychopathology. METHODS To add clarity, the present study utilized an unusually large, national sample, all subjects of which were administered a standardized, comprehensive test of psychopathology (the MMPI-2), an intensive interview concerning psychosocial history, and a medical examination. Subjects' scores on the MMPI-2 were compared to available norms. The psychosocial interview yielded information about families of both origin and reference. Information about comorbidities, medications, and body mass index (BMI) were available from the medical examinations. Multiple regression analyses were performed to determine the variables that best predicted psychopathology and BMI. RESULTS The percentage of subjects whose MMPI-2 scores exceeded and approached psychopathological levels was in excess of normative expectations. The 1,3,2 pattern of scale scores on this test expresses depressive disorder, with anxiety and somatization features. Regression analyses showed that abuse of the subject and of substances in the family of origin positively predicted, while education and number of children negatively predicted both psychopathology and BMI. CONCLUSIONS Dysfunctionality in the family of origin may lead to obesity through such regressive coping styles as stress eating, but this can be offset by personal development and nurturance responsibilities.