BACKGROUND: Barrett's esophagus is a strong risk factor for esophageal adenocarcinoma, but little is known about its associations with body mass index (BMI) or abdominal obesity. METHODS: We conducted a case-control study within the Kaiser Permanente Northern California population. Persons with a new diagnosis of Barrett's esophagus (cases) were matched to subjects with gastroesophageal reflux disease (GERD) without Barrett's esophagus and to population controls. Subjects completed questionnaires and an anthropometric examination. RESULTS: We interviewed 320 cases, 316 patients with GERD, and 317 controls. There was a general association between Barrett's esophagus and a larger abdominal circumference (independent of BMI) compared with population controls (odds ratio, 2.24; 95% confidence interval, 1.21-4.15; circumference, >80 cm vs <80 cm). There was a possible risk plateau, with increased risk evident only at circumferences >80 cm and no significant trend for further increases in circumference. There was a trend for association compared with patients with GERD (test for trend, P = .03). There was no association between Barrett's esophagus and BMI. Abdominal circumference was associated with GERD symptom severity (odds ratio, 1.86; 95% confidence interval, 1.03-3.38; risk of severe weekly GERD, per 10-cm circumference); adjustment for GERD partially attenuated the association between Barrett's esophagus and circumference. CONCLUSIONS: Waist circumference, but not BMI, had some modest independent associations with the risk of Barrett's esophagus. The findings provide partial support for the hypothesis that abdominal obesity contributes to GERD, which may in turn increase the risk of Barrett's esophagus.