To compare open-magnet magnetic resonance (MR) imaging performed with the patient sitting with dynamic closed-magnet MR imaging of the pelvic floor performed with the patient supine. Thirty-eight patients underwent dynamic 1.5-T closed-magnet pelvic floor MR imaging while in the supine position. Midsagittal T2-weighted single-shot fast spin-echo and T1-weighted multiphase spoiled gradient-recalled-echo (SPGR) MR images were obtained before and after rectal contrast agent administration, respectively, with the patient at rest, straining, and maximally contracting the sphincter. Subsequently, the patient was transferred to an open 0.5-T system. Midsagittal multiphase T1-weighted SPGR MR images were then obtained every 2 seconds with the patient sitting while at rest, maximally contracting the sphincter, straining, and defecating. Images were analyzed with regard to presence of enteroceles, anterior rectoceles, intussusceptions, rectal descents, bladder descents, and vaginal vault descents. All intussusceptions were missed at supine MR imaging. With sitting MR imaging as the reference standard, the sensitivity of supine MR imaging was 79% for depiction of bladder descents. When MR findings were graded and clinically irrelevant MR findings were excluded, sensitivity increased to 100% for depiction of bladder descents and anterior rectoceles and to 96% for depiction of rectal descents. Dynamic supine MR imaging performed with a closed-configuration unit before and after rectal contrast agent administration appears to be an alternative to sitting MR defecography performed with an open-configuration unit for diagnosis of clinically relevant pelvic floor abnormalities. Copyright RSNA, 2002