In pts with IDA with clinical suspicion for gastrointestinal blood loss without localizing symptoms, the standard practice is to perform a colonoscopy (C) followed by an EGD, if C is negative. If neither endoscopy reveals an etiology, the small bowel is subsequently evaluated, generally with CE. Objective: To prospectively determine the yield and cost-effectiveness of CE as the initial diagnostic test in pts with IDA who have a negative C. Methods: All pts 18 yrs and older who were diagnosed with IDA using strict criteria and referred for endoscopic evaluation were prospectively enrolled. Pts were excluded if they had contraindications for CE, had menometrorrhagia, previous endoscopic evaluation, prior surgical resection of small bowel and/or stomach, and prior GI malignancies. Patients with negative C underwent CE followed by EGD within 72 hrs. Both the capsule reader and endoscopist were blinded to the alternate test. Pts were prospectively followed up to one year. Results: 38 pts (20 men; 18 women; median age 56; range 44-81) with IDA were referred for endoscopic evaluation; 12 pts (32%) had a positive C. 6 pts declined to participate. CE revealed a definitive source for IDA in 12 pts (5 arteriovenous malformations, 3 NSAID ulcerations and/or strictures, 1 gastric ulcer, 1 severe erosive gastritis, 1 celiac sprue, 1 jejunal adenocarcinoma) and a probable source in 4 pts (2 NSAID erosions, 1 erosive gastritis, 1 jejunal diverticulosis). EGD revealed a definitive source in 4 pts (1 gastric ulcer, 1 severe erosive gastritis, 1 celiac sprue, 1 Cameron's erosions in a hiatal hernia sac) and a probable source in 1 pt (erosive gastritis). The yield of CE compared to EGD was 80% versus 25%. EGD identified only one patient (Cameron's erosions) with a source for IDA not identified by CE. Performing CE first in patients with IDA and a negative C resulted in an average cost savings of $386 per pt. Prospective follow-up of all pts revealed that 5 pts had additional testing done; one additional pt with arteriovenous malformations was noted on repeat CE. Conclusions: This is the first prospective study to demonstrate that performing CE initially in pts with IDA and a negative C results in a more cost-effective yield. Additional data are required before considering a paradigm shift in the standard practice of endoscopic evaluation in pts with IDA who have a negative C.