Polymerase chain reaction (PCR) for the diagnosis of Clostridium difficile infection (CDI) might result in overdiagnosis. The clinical outcomes of symptomatic CDI patients diagnosed by PCR remain uncertain. We aimed to determine whether patients whose diagnosis of CDI was based on PCR had different characteristics and clinical outcomes than those diagnosed by toxin immunoassay. Consecutive CDI patients, hospitalized at Rabin Medical Center, Beilinson Hospital, Petah Tikva, Israel, between January 2013 and January 2016, were identified retrospectively and included in the study. Diagnosis of CDI was based on PCR or diagnosis by immunoassay for C . difficile toxin. The main outcome was 30- and 90-day all-cause mortality. The PCR group included 165 patients and the immunoassay group included 157 patients. In comparison to the immunoassay group, patients in the PCR group were more likely to be younger, to be independent, to undergo previous abdominal surgery, and to use laxatives. The 30-day mortality rate in the PCR group was significantly lower than that in the immunoassay group, 29/165 (18%) vs 49/157 (31%), respectively; p  = 0.028. On multivariate analysis, PCR diagnosis was associated with reduced mortality, OR 0.48 (95% CI 0.26–0.88). PCR-based diagnosis of CDI is associated with reduced all-cause mortality rates. Further studies are needed to determine the management of patients with discrepant immunoassay and PCR diagnosis of CDI.