A computerized health records system has facilitated an analysis of the correlated illness of a segment of population afflicted with a chronic disease. Taking the existence of clinical diabetes as a starting point, the total illness experience of 635 diabetics, as shown by clinic utilization data, has been compared with an age-sex-geographically matched control population.The comparisons show that, overall, diabetics cause a caseload rate more than double that of nondiabetics. In 17 out of 22 broad health problem areas, the diagnosis caseload rates for diabetics varied from 1.3 to 3.7 times greater than the rate for matched, non-diabetic controls. Workload (physician visit and hospital days) comparisons showed even greater disparity.Some of the comparisons confirm statistically the clinical impressions of diabetologists over the past years. Some of the findings are not reflections of orthodox clinical opinion. The results contribute to further elucidation of the natural history of this disease. They also indicate the need for the development of comprehensive clinical management programs for groups of interrelated conditions found in persons with chronic disease such as diabetes.