Purpose: To determine the frequency, type and significance of UGI findings found on antegrade deep enteroscopy (Double Balloon (DBE) and Spiral) performed for evaluation of obscure GI bleeding at a referral center after EGD, colonoscopy and capsule endoscopy for patient selection. Methods: Patients referred to our center who underwent antegrade deep enteroscopy for obscure GI bleeding were included. All had undergone EGD and colonoscopy and all had capsule studies re-read, with findings warranting deep enteroscopy. Per our protocol, antegrade DBE and Spiral enteroscopy were done under GA, avoiding narcotics, and after transfusion to Hgb 10. Data analyzed included patient demographics, type and extent of exam, overall findings in upper GI territories (gastric, D1, D2, D3, D4 and ligament of Treitz) and type of finding (angioectasia, ulcers, GAVE, other). Clinical significance of UGI findings was judged based on type of lesions, number and size of UGI lesions relative to distal findings and presence of active bleeding. Results: One hundred nineteen deep enteroscopy procedures (110 DBE and 9 Spiral) for obscure GI bleeding indications were performed on 119 patients (Mean age 66.8 years; 50 M, 69 F). Mean depth of insertion was estimated at 250 cm beyond the pylorus (range 60 to 550). Overall, 73% (87/119) of deep enteroscopy exams had significant lesions found. Notably, 46% (55/119) had UGI lesions found well within push enteroscopy territory (gastric, D1, D2, D3, D4 and ligament of Treitz) with 31% (37/119) of lesions within typical EGD territory (G, D1 and D2). 75% (41/55) of exams with UGI lesions had angioectasias only, while 25% (14/55) had other lesions including GAVE (3), ulcers (6) and other (5). Clinical significance of these UGI lesions was judged as: Predominant source of bleeding 10, significant contributor 25, possible contributor 15, not significant 4. Conclusion: In our series, 46% of antegrade deep enteroscopy exams had lesions found within reach of push enteroscopy and 31% had lesions within reach of standard EGD. These UGI lesions, mostly angioectasias, were judged to be a predominant or significant source of GI bleeding in 29% of exams overall (35/119). We consider that this relatively high frequency of UGI findings may be in part due to improved detection of angioectasias by avoidance of narcotics and transfusion to Hgb 10. Our data also reinforces the value of âsecond lookâ EGD or push enteroscopy in evaluation of obscure GI bleeding. Disclosure: Dr Brown, Speakers Bureau: Given Imaging (Capsule Endoscopy).