Objectives Clostridium difficile infection (CDI) is a common complication in inflammatory bowel disease (IBD) and has been associated with poor IBD outcome. The aims of our study were to look for predictive factors of CDI in patients hospitalized for IBD flare and to evaluate a rapid testing strategy in this population. Methods Consecutive patients hospitalized for IBD flare in Saint-Antoine Hospital (Paris, France) were prospectively tested for CDI with a defined strategy involving rapid testing and reference methods. Risk factors for CDI were investigated and performances of diagnostic tests were evaluated. Results C. difficile testing was performed at admission in 461 hospitalizations for IBD flare. CDI was diagnosed in 35 cases (7.6%) and non-toxigenic C. difficile was identified in 10 cases (2.2%). In multivariate analysis, UC phenotype was associated with CDI (OR 2.2, 95% CI 1.03–4.6, p = 0.047). Glutamate dehydrogenase (GDH) test had a 97.1% sensitivity and a 100% negative predictive value for CDI diagnosis but a positive predictive value of 79.1%. Enzyme immunoassay (EIA)-based toxin detection (C. Diff Quik Chek complete®, Alere) had a poor sensitivity and diagnosis was rescued by toxin PCR in 100% of cases. Conclusion CDI is frequent in patients hospitalized for IBD flare. Clinical parameters do not help for the diagnosis and rapid testing should be performed in all patients. Currently, a negative result of an EIA-based toxin search associated with a positive GDH test cannot rule out a CDI and should not delay initiation of specific treatment in case of severe symptoms or high presumption.