A rapid (2.5 h) direct enzyme immunoassay (EIA) for Clostridium difficile toxin A was developed for clinical use. Specimen centrifugation and filtration were not required. The EIA detected toxin A levels in patient stool as low as 20 pg (2 ng/ml of stool). The test was 5,000 times more sensitive for toxin A than it was for toxin B and did not react with a panel of other bacterial species with the exception of one highly toxigenic strain of Clostridium sordellii. The EIA was compared with the cytotoxin assay, culture of toxigenic C. difficile (toxigenic culture), and latex agglutination by using 313 fresh stool specimens submitted from patients with suspected C. difficile-associated disease. Results read visually and with a plate reader were similar. Sixty-two specimens were positive by one or more tests, but only 22 (35%) were positive by all four laboratory methods. The EIA was 84.1% sensitive and 98.9% specific when it was compared with the cytotoxin assay. The use of toxigenic culture to referee discrepant results (EIA versus cytotoxin assay) showed the EIA sensitivity and specificity to be 95.1 and 99.3%, respectively, with respect to other laboratory methods. Patient charts were reviewed for antibiotic-associated diarrhea on 108 specimens, including all those that were positive by at least one test method. Of 34 patients determined to have C. difficile-associated disease, 29 (85.3%) were positive by EIA, 32 (94.1%) were positive by the cytotoxin assay, 27 (79.4%) were positive by toxigenic culture, and 20 (58.8%) were positive by latex agglutination. Seven patients with antibiotic-associated diarrhea had a positive latex result, but results were negative by EIA, the cytotoxin assay, and toxigenic culture. The EIA demonstrated high specificity and good sensitivity for C. difficile-associated disease cases. The test can be used alone or in combination with the cytotoxin assay or toxigenic culture to provide rapid and sensitive results.