This study was carried out in order to investigate the frequency of Clostridium difficile toxin in the stool samples of hospitalized 74 neutropenic (mean age: 41.78 +/- 14.3 years; 40 male) and 75 non-neutropenic patients (mean age: 44.09 +/- 15.6 years; 30 male) who developed antibiotic-associated diarrhea between January 2003-September 2004 in a university hospital and also to analyze the related risk factors. C. difficile toxin A and toxin A/B were searched by immunochromatographic method (Toxin Detection Kit, Oxoid, UK), and commercial ELISA (Clostridium difficile Antigen, Generic Assays GmbH, Germany) kit, respectively. Stool samples were also analysed in terms of the presence of other bacterial and parasitic agents which may cause diarrhea. Statistical evaluation were performed by Kaplan-Meier survival analysis and by Cox regression analysis. Both neutropenic and non-neutropenic groups were compared according to their incidence densities based on times in days of overall hospitalization, total antibiotic use and hospitalization until diagnosis. The antibiotics used in neutropenic patients were piperacillin-tazobactam (41.9%), imipenem (25.7%), cefepime (17.5%), ciprofloxacin (2.7%) and others (12.2%) and in non-neutropenic patients were ampicillin-sulbactam (29.3%), ciprofloxacin (18.7%), ceftriaxone (14.6%) and others (30.7%). C. difficile toxin A positivity rates in neutropenic and non-neutropenic groups were found as 13.5% (10/74) and 14.7% (11/75), respectively, with a total rate of 14.1% (21/149). The positivity rate of toxin A/B was 24.3% (n= 18) for neutropenic, and 21.3% (n= 16) for non-neutropenic patients, with a total rate of 22.8% (n= 34). There was no statistically significant difference between the groups by means of toxin A or toxin A/B positivity rates (p > 0.05). Thirteen (38.2%) of 34 toxin A/B positive patients yielded negative results with toxin A detection test. Our results revealed that infection in the neutropenic patients developed much earlier than that in the non-neutropenic group by comparing durations of hospitalization and antibotic use which were shorter for toxin positive individuals (p < 0.01 and p < 0.001, respectively). In the control group, implementation of sulbactam-ampicillin or amoxicillin-clavulanate was determined as a risk factor. In addition to duration of hospitalization, use of antibiotics was evaluated as a risk factor for C. difficile associated colitis, especially in the neutropenic group. According to these results, it is possible to point out that antibiotic-associated colitis develops relatively earlier in neutropenic patients and is more frequently C. difficile toxin positive. In this context, appropriate control measures should therefore be kept in mind.