Purpose: Improvements in tolerogenic immunosuppressive therapy have resulted in major improvements in small bowel transplantation (SBTx) survival during the past decade, such that survival parallels that for liver transplant recipients: >90% for the first year. Although rejection may be more frequent with SBTx because of the immunogenic character of the intestine, it is easier to detect early acute cellular rejection (ACR) due to endoscopy, and if detected early, ACR is rapidly reversible with steroid bolus therapy. In the 1st yr following SBTx, a short segment of ileum is interconnected between the colon, ileal graft and abdominal wall that makes graft access easy. The objective of this study was to compare visual findings to histology from chimney biopsies in asymptomatic and to more invasive ileal and jejunal biopsies in symptomatic patients. Methods: We analyzed the single experience of an experienced endoscopist over 12 months. 590 surveillance graft ileoscopies and biopsies were performed in 67 stable asymptomatic patients (Table 1), and 40 combined ileoscopies and enteroscopies in 19 symptomatic patients. The surveillance microscopic findings were divided into 3 subcategories: mucosal architecture (normal vs. abnormal: e.g. distortion, ulceration, regeneration, granulation, atrophy), lamina propria inflammation (normal vs. abnormal: e.g. edema, congestion, hemorrhage, inflammatory infiltrate), and crypt apoptosis (normal <0–4 apoptotic bodies/10 HPF, abnormal ≥5).Table: Surveillance endoscopies.Results: In surveillance, the positive predictive value of visualization for normal histology was 83.5% for mucosal architecture, 34.3% for lamina propria, and 94.5% for apoptosis score. The predictive value of visualization for abnormal histology was 77.5% for mucosal architecture, 93.8% for lamina propria, and 33.8% for apoptosis score. The correlation between visualization and the microscopic findings was significant overall for the 3 variables (Pearson Chi-Square; P < 0.01). In symptomatic patients, ACR was detected in 24/40 endoscopies although the mucosa looked normal in 14. In 30 pts where chimney and ileal biopsies were taken, ACR was detected in 21, with 16 showing ACR in both sites and 4 ACR in the chimney alone. In 10 patients jejunal biopsies were taken in addition to chimney showing the same rate of ACR in both sites. This suggests a sensitivity of 94% for chimney biopsy detection of graft ACR. Conclusion: Although endoscopic visualization identifies most patients with abnormal histology, early features of ACR may be missed, making biopsy mandatory, even in high risk coagulopathic patients. The high sensitivity of chimney biopsy in detecting ACR endorses the priority of early, less invasive, chimney biopsy in the surveillance of SBTx patients.