Objectives: In human immunodeficiency virus (HIV)-infected patients with chronic unexplained diarrhea, upper endoscopy with small bowel biopsy and aspirate is often performed to identify treatable pathogens. The purpose of this study was to compare the diagnostic yield of duodenal with jejunal biopsy and aspirate. Methods: All HIV-infected patients with chronic unexplained diarrhea who were evaluated by upper endoscopy at Bellevue Hospital Center between January 1992 and January 1997 were identified. Data were collected by reviewing patient charts, endoscopy reports, and pathology records. Results: During the 5-yr study period, 442 patients underwent upper endoscopy with sampling of the duodenum (N = 173) or jejunum (N = 269). A pathogen was identified in 123 patients (27.8%). Microsporidia was the most common organism detected (12.2%). The diagnostic yield of jejunal biopsy and aspirate was significantly higher than that obtained from the duodenum (32.3% vs 20.8%, p = 0.009). Small bowel aspirates detected a pathogen in only 1.8% of patients evaluated, and there was no difference in the yield of duodenal and jejunal aspirates (1.3% vs 2.1%, p = 0.7). Patients with a CD4 count of < 100 cells/mm3 were significantly more likely to have a pathogen identified than those with higher CD4 counts (38.8% vs 7.1%, p < 0.0001). Conclusions: Upper endoscopy with small bowen biopsy and aspirate identifies a pathogen in 27.8% of individuals with HIV-related chronic unexplained diarrhea. In this patient population, jejunal biopsies acquired by enteroscopy are superior to those obtained from the duodenum. Small bowel aspirates are of little value in the workup of chronic HIV-related diarrhea.