To describe the extent of intentional insulin omission in an outpatient population of women with insulin-dependent diabetes mellitus (IDDM) and examine its relationship to disordered eating, attitudes toward diabetes, other psychosocial factors, long-term complications, and glycemic control. Before their routinely scheduled clinic appointments, female IDDM patients who were 13-60 years of age completed a self-report survey (final n = 341). The survey included standardized questionnaires assessing disordered eating attitudes and behaviors, psychological functioning (general distress, diabetes-specific distress, and hypoglycemic fear), attitudes toward diabetes, and self-care behaviors. All subjects were assessed for glycosylated hemoglobin within 30 days of survey completion. Long-term complications were determined through chart review. Approximately 31% of the subject sample, representing women of all ages, reported intentional insulin omission, but only 8.8% reported frequent omission. Compared with non-omitters, omitters reported more disordered eating, greater psychological distress (general and diabetes-specific), more hypoglycemic fear, poorer regimen adherence, and greater fears concerning improved diabetes management (which may lead to weight gain). Omitters evidenced poorer glycemic control, more diabetes-related hospitalizations, and higher rates of retinopathy and neuropathy. Multivariate examination revealed only two variables that independently predicted omission: diabetes-specific distress and fear of improved glycemic control ("because I will gain weight"). Of the omitters, approximately half reported omitting insulin for weight-management purposes (weight-related omitters). These subjects evidenced significantly greater psychological distress, poorer regimen adherence (including more frequent omission), poorer glycemic control, and higher rates of complications than did non-weight-related omitters as well as non-omitters. Non-weight-related omitters tended to fall between weight-related omitters and non-omitters on most measures of psychological functioning, adherence, and glycemic control. These findings suggest that insulin omission is common, that it is not limited to younger women, and that the medical consequences of omission, especially frequent omission, may be severe. Although a strong association between omission and disordered eating was observed, these data suggest that this link may be complicated by important diabetes-specific factors. Patients preoccupied with eating and weight concerns may also become emotionally overwhelmed by diabetes and/or fearful of normoglycemia (and the associated weight-related consequences), thus reinforcing the desire to omit insulin and maintain elevated blood glucose levels.