OBJECTIVE--To evaluate attitudes and practices of primary-care physicians toward tight blood glucose control in IDDM. RESEARCH DESIGN AND METHODS--A mail and telephone questionnaire survey was conducted on a systematic, stratified sample of 1429 family-practice physicians, general practitioners, internists, and pediatricians in active practice in the United States who treated patients with IDDM. Physicians were asked about methods they used for clinical and laboratory assessment of blood glucose control and about their attitudes and beliefs in treating IDDM. They were asked also what they consider to be acceptable ranges for blood glucose and HbA1 in IDDM patients. A score was developed reflecting three criteria for tight blood glucose control: fasting glucose 70-120 mg/dl (3.9-6.7 mM), 2-h postprandial glucose < 180 mg/dl (< 10 mM), and HbA1 < or = 8% (the nondiabetic value was specified as 5-7%). Physicians were accorded one point when their acceptable range agreed with an intensive treatment criterion (range for score 0-3). RESULTS--Only 31% of physicians agreed with all three criteria for tight control of blood glucose; 37% agreed with none or only one of the standards. Pediatricians were particularly low in their agreement with the HbA1 standard. Physicians who agreed with one of the three criteria often did not agree with the other two. With increasing value for the score, there was a greater proportion of physicians whose management practices (e.g., frequent measurement of HbA1, multiple insulin injections, patient SMBG, use of dietitian/educator in care of patients) are conducive toward tight control of blood glucose. However, even among physicians with a score of 3, HbA1 was ordered infrequently, three or more insulin injections/day was prescribed rarely, patient SMBG was less than fully endorsed, and both a dietitian and diabetes educator were used by a minority of physicians. CONCLUSIONS--It appears that primary-care physicians are not fully aware of recommended criteria for intensive treatment of blood glucose in IDDM patients or of the importance of multiple insulin injections, use of HbA1, and patient SMBG. Physician practice behaviors are less than optimal for intensive management of IDDM patients, even among physicians who agree with all three standards for intensive treatment of blood glucose in IDDM.