Recent epidemiological studies suggested that cpb2-positive Clostridium perfringens isolates are associated with gastrointestinal (GI) diseases in horses. These putative relationships, indicated by PCR genotyping, were tested in the present study by further genotyping and phenotyping of 23 cpb2-positive C. perfringens isolates from horses with GI disease (referred to hereafter as horse GI disease isolates). Our beta2-toxin (CPB2) Western blot analyses demonstrated that all of the tested isolates were unable to produce detectable levels of CPB2. However, Southern blot and nucleotide sequencing analyses identified intact cpb2 open reading frames in all of our surveyed horse GI disease isolates. Furthermore, reverse transcriptase PCR and Northern blot analyses showed that cpb2 genes in all of our surveyed horse GI disease isolates were transcriptionally active, i.e., an approximately 1.2-kb cpb2-specific mRNA was identified in total RNA from our surveyed isolates. The levels of cpb2 mRNA in CWC245 (a high-CPB2-producing pig strain) and our surveyed horse GI disease isolates differed to such an extent (35-fold) that this difference could be considered as a major cause of the difference in levels of CPB2 production by CWC245 and horse GI disease isolates. This finding received further support from our observation that the complementing strain 106902(pMRS140), which produced significantly higher levels of mRNA than strain 106902, produced high levels of CPB2. Collectively, our results indicated that there is a positive correlation between cpb2 transcription levels and the amount of CPB2 produced by a C. perfringens cell and that decreased transcription and/or message instability may be involved, at least in part, in the low CPB2 production noted for horse GI disease isolates in comparison to that noted for pig GI disease isolate CWC245.