In this study, clinical and laboratory findings were tested for correlation with the presence of Clostridium difficile. The toxigenicity of the isolated strains and the toxins were determined in faecal samples of immunocompromised children admitted to a single room for protective isolation. Using the toxin assay as the gold standard, the culture sensitivity of toxigenic C difficile was 94.1%, the specificity 93.8%, the positive predictive value 62.8%, and the negative predictive value 99.3%. Correction for stools with a positive culture of toxigenic C difficile preceding detection of toxin, resulted in a positive prediction value of 78.4%. A statistically significant association was found between a positive faecal toxin assay and fever, and between a positive culture of toxigenic C difficile and abdominal pain: 42% of the patients with positive toxin assays had fever versus 21% with negative toxin assays, and 66% of the patients with a positive culture for toxigenic C difficile had abdominal pain, versus 22% with negative cultures. Further analysis of the cultures and toxin assays showed no statistically significant association with diarrhoea, fever, white blood cell count, C reactive protein concentrations, or abdominal pain. Based on these findings, it is suggested that immunocompromised children should be treated when toxigenic C difficile is cultured or when toxin is detected in stool samples.