In 1990, the Division of Endocrinology and Metabolism of Henry Ford Hospital established an Outcomes Management data base for patients with Type I and Type II diabetes. A first cohort of 117 patients completed a baseline and 6-month follow-up assessment; a second cohort of 116 patients completed the baseline assessment. Assessment at each time point includes: the Short Form--36 Questions (SF-36) health status instrument; a set of clinical variables known as the Diabetes TyPE scale Form 2.2 abstracted from the medical record; and the physicians' ratings of patient's health status along the major dimensions of the SF-36. Success with both face-to-face and mailed administration of the SF-36 has been good, with response rates of over 85% using both methods. Comparison of patient and physician ratings of patient health status indicated a significant discrepancy on ratings of general health status, with physicians' ratings higher than those of patients themselves. "Tight" glycemic control (as measured by glycosylated hemoglobin) was associated with somewhat lower ratings on the various SF-36 dimensions for all patients in the first cohort and for Type I patients in the second cohort. However, this effect did not seem to be attributable to those features of a complex regimen used to achieve tight control, but rather reflected a complex combination of age, education level, and number of daily injections associated with achieving good control.