OBJECTIVE The molecular epidemiology of pediatric Clostridium difficile infection (CDI) is poorly understood. We aimed to identify the restriction endonuclease analysis (REA) groups causing CDI and to determine risk factors and outcomes associated with CDI caused by epidemic strains in children. DESIGN Retrospective cohort study PATIENTS Inpatients and outpatients >1 year old receiving care between December 2012 and December 2013 SETTING An academic children's hospital in Chicago, Illinois METHODS C. difficile PCR-positive stools were cultured, and C. difficile isolates were typed by REA. REA of isolates from patients with multiple CDIs was performed to differentiate relapse (infection with same strain) from reinfection (different strains) irrespective of time between CDIs. RESULTS A total of 189 CDIs occurred among 145 patients. REA groups were widely distributed. The BI/NAP1/027 strain caused CDI in only 1 patient. DH/NAP11/106, the predominant epidemic strain identified, was associated with the use of third- or fourth-generation cephalosporins (risk ratio [RR], 3.2; 95% confidence interval [CI], 1.1-9.9; P=.04). CDI relapse commonly occurred up to 20 weeks later. Compared with CDI caused by non-DH/NAP11/106 strains, CDI caused by DH/NAP11/106 was more likely to result in multiple CDI relapses (40% vs 8%; P=.05) among children with multiple CDIs. CONCLUSIONS REA identified the exceedingly low prevalence of BI/NAP1/027 and the high prevalence of DH/NAP11/106, a common epidemic strain in the United Kingdom that is less often reported in the United States. CDI relapse commonly occurred up to 20 weeks from the previous CDI. Defining recurrent CDI as that occurring only within 8 weeks of the original infection may lead to misclassification of some recurrent CDIs as new CDIs in children. Infect Control Hosp Epidemiol 2015;00(0): 1-7.