BACKGROUND: Optimal management of obscure gastrointestinal bleeding (OGIB) remains unclear.OBJECTIVE: To evaluate diagnostic yields and downstream clinical outcomes comparing video capsule endoscopy (VCE) with push enteroscopy (PE).METHODS: Patients with OGIB and negative esophagogastroduodenoscopies and colonoscopies were randomly assigned to VCE or PE and followed for 12 months. End points included diagnostic yield, acute or chronic bleeding, health resource utilization and crossovers.RESULTS: Data from 79 patients were analyzed (VCE n=40; PE n=39; 82.3% overt OGIB). VCE had greater diagnostic yield (72.5% versus 48.7%; P<0.05), especially in the distal small bowel (58% versus 13%; P<0.01). More VCE-identified lesions were rated possible or certain causes of bleeding (79.3% versus 35.0%; P<0.05). During follow-up, there were no differences in the rates of ongoing bleeding (acute [40.0% versus 38.5%; P not significant], chronic [32.5% versus 45.6%; P not significant]), nor in health resource utilization. Fewer VCE-first patients crossed over due to ongoing bleeding (22.5% versus 48.7%; P<0.05).CONCLUSIONS: A VCE-first approach had a significant diagnostic advantage over PE-first in patients with OGIB, especially with regard to detecting small bowel lesions, affecting clinical certainty and subsequent further small bowel investigations, with no subsequent differences in bleeding or resource utilization outcomes in follow-up. These findings question the clinical relevance of many of the discovered endoscopic lesions or the ability to treat most of these effectively over time. Improved prognostication of both patient characteristics and endoscopic lesion appearance with regard to bleeding behaviour, coupled with the impact of therapeutic deep enteroscopy, is now required using adapted, high-quality study methodologies.