The indications for MR cholangiopancreatography (MRCP) in children, and its safety and findings, might differ from those in adults and are not well described. To investigate the safety, feasibility, and accuracy of MRCP in children. We reviewed all prospective MRCP reports, noting the indication, the use of secretin, endoscopic retrograde cholangiopancreatography (ERCP) findings, and patient outcomes. Two readers reviewed each MRCP study by consensus to rate duct visualization and compare pancreatic duct sizes before and after secretin administration (paired t-test). The likelihood of a normal versus an abnormal MRCP study was compared by gender, pancreatitis as the primary indication, secretin use, and whether ERCP was performed (Fisher's exact test), as well as age (t-test). A total of 85 MRCP studies were performed in children (mean age 10.3 years), most commonly for evaluation of pancreatitis (n=47, 55%); 41 (48%) used secretin and 39 (46%) used a negative oral contrast agent. Of the 85 studies, 72 (85%) had excellent image quality and 43 were normal. ERCP was performed after 16 of the 85 MRCP studies (19%); the diagnoses were concordant with those of MRCP in 13 (81%). There were 42 abnormal MRCP studies, and these were more likely to be in girls (P=0.03) and in children who had undergone ERCP (P<0.01). Secretin and the negative oral contrast agent were well-tolerated. Secretin improved duct visualization (P<0.001). MRCP safely and accurately depicted pancreaticobiliary anatomy in children. The use of secretin improved visualization of the pancreatic duct.