It is estimated that obesity has clinical and economic repercussions comparable to those of AIDS or cancer. To learn more about the opinions and attitudes of obstetricians-gynecologists concerning their obese patients, the authors conducted a survey of the fellows and junior fellows of the American College of Obstetricians and Gynecologists (ACOG). The survey contained questions about weight management and assessment practices, the importance of weight control. to patient health, disease risks associated with obesity, and self-assessment of their own preparation for dealing with problems of weight management. More than 1100 surveys were sent to ACOG members. Five hundred twenty-five (47%) surveys were returned and analyzed. Most of the respondents (79%) practiced both obstetrics and gynecology in suburban (30%) or urban (28%) settings. The physicians reported a patient population that was, on average, 64% white, 15% African American, 14% Hispanic, 4% Asian, and 2% Native American. Nonpregnant patients were weighed at every visit in 60% of the practices and nearly every visit in an additional 34%. Ten percent of the respondents calculated the body mass index (BMI) for all of their nonpregnant patients; 20% indicated that they would determine a BMI only if the patient was concerned about her weight. Visual inspection was the most frequently indicated of the possible assessments of obesity (64%), but 56% of respondents reported using (either in addition to other methods or alone) a BMI of greater than 30 kg/m2 as the determinant of obesity. The physicians were asked to estimate the percentage of patients in their practices who were underweight, overweight but not obese, and obese. On average, the respondents reported 6% underweight, 36% overweight but not obese, and 21% obese patients. There was a positive correlation between the percentage of the patients in a practice who were Hispanic or African American and the prevalence of obesity in the practice (P <.001). Conversely, the percentage of obese patients was negatively correlated with the percentage of non-Hispanic white or Asian women in a practice (P =.003). Eight-two percent of the respondents said that weight management was an important (66%) or very important (16%) part of their practice, and nearly half indicated that they thought that this practice component would increase in importance in the next few years. Forty-nine percent of the respondents thought that a patient's weight was very important component of her overall health. Weight was rated a very important factor in the management of pregnant patients by 54%, in the management of hypertension by 76%, and in the management of diabetes by 84%. Almost 80% had referred patients to other professionals for weight management, and 36% had prescribed some form of weight loss medication to their patients. A large majority of the respondents agreed that obesity is a major health problem in the United States and in their own practices, the benefits of weight loss outweigh the risks (for nonpregnant patients), and obesity is an indicator of an unhealthy lifestyle. They also thought that weight reduction can improve health, that the health risks of obesity are real and not overstated, and that obesity is a condition that should be treated. Most responding physicians knew that obesity increases the risk of diabetes, cardiovascular disease, hypercholesteremia, hyperinsulinemia, menstrual irregularities, infertility, preeclampsia,, depression, and colon and breast cancer. They did not think that obesity was a factor in osteoporosis, ovarian cancer, preterm labor, lung cancer, endometriosis, intrauterine growth retardation, premenstrual syndrome, or birth defects. The respondents indicated that their medical school or residency training included little or no information on the risks of obesity or weight management for patients. They reported that their knowledge of obesity and weight management came from reading journals, ACOG