Clostridium perfringens enterotoxin (CPE) is an important sporulation-associated virulence factor in several illnesses of humans and domestic animals, including C. perfringens type A food poisoning. Therefore, the ability to determine the enterotoxigenicity of food or fecal C. perfringens isolates with simple, rapid assays should be helpful for epidemiologic investigations. In this study, Western immunoblotting (to detect CPE production in vitro) was compared with PCR assays and digoxigenin-labeled probe assays (to detect all or part of the cpe gene) as a method for determining the enterotoxigenicity of C. perfringens isolates. The cpe detection assays yielded reliable results with DNA purified from vegetative C. perfringens cultures, while Western immunoblots required in vitro sporulation of C. perfringens isolates to detect CPE production. Several cpe-positive C. perfringens isolates from diarrheic animals did not sporulate in vitro under commonly used sporulation-inducing conditions and consequently tested CPE negative. This result indicates that cpe gene detection and serologic CPE assays do not necessarily yield similar conclusions about the enterotoxigenicity of a C. perfringens isolate. Until further studies resolve whether these cpe-positive isolates which do not sporulate in vitro can or cannot sporulate and produce CPE in vivo, it may be preferable to use cpe detection assays for evaluating C. perfringens isolate enterotoxigenicity and thereby avoid potential false-negative conclusions which may occur with serologic assays.