Diarrhea is one of the dose-limiting toxicities for administration of fluorouracil (5FU) in patients with colorectal cancer and can result in severe morbidity and mortality. No well-defined prognostic factors influencing 5FU-associated diarrhea have been identified, which means its occurrence is unforeseeable. The aim of this study was to check whether any characteristics related to patients or chemotherapy could allow the identification of subsets of patients at higher risk of developing diarrhea while receiving a regimen containing 5FU. A logistic regression analysis was performed with age, sex, site of primary tumor, presence of primary tumor, presence of colostomy, time since surgery, number of courses of chemotherapy, diarrhea in previous courses, season of treatment, and chemotherapeutic regimens used as model parameters to predict occurrence of diarrhea in 258 colorectal cancer patients receiving a 5FU-containing regimen. Presence of primary tumor (P = 0.004), previous episodes of chemotherapy-related diarrhea (P = 0.00005) and summer season (P = 0.014) were found to be significant risk factors for developing diarrhea. The other variables examined, such as age, sex, chemotherapeutic regimen, site of primary tumor, presence of colostomy, and time since surgery, were not significantly correlated to diarrhea. Chemotherapeutic regimen was the only parameter that allowed prediction of the severity of diarrhea: 5FU/6S-leucovorin/interferon caused more severe diarrhea, followed by 5FU/leucovorin weekly. Although the analysis of these clinical features does not seem to allow the definition of a well-defined subset of colorectal cancer patients at higher risk of 5FU-induced diarrhea, it can be recommended that patients with primary tumor, or who have experienced diarrhea in earlier courses of chemotherapy or are receiving treatment in summer should be carefully monitored, especially in the first cycles.