We aimed to assess technical feasibility, clinical applicability, and diagnostic yield of videocapsule endoscopy (VCE) in a large group of unselected patients. VCE was performed with the Given Imaging swallowable capsule. Findings were considered diagnostic if the observed finding could explain the symptomatology of the patient. Findings were considered suspicious if an observed finding failed to completely explain the patient's symptoms. We studied 250 patients. A definite diagnosis was made in 95 patients (38%). Suspicious findings were noted in 80 patients (32%). No diagnosis was obtained in 74 patients (30%). The yield of VCE was higher in patients with suspected Crohn's disease. Mean viewing time decreased significantly from 51 +/- 14 to 30 +/- 7 minutes after reviewing 50 procedures. VCE is an important diagnostic tool, but a definite diagnosis is established in only 38% of the patients. The highest diagnostic yield is obtained in patients with suspected Crohn's disease.