We describe here the use of an immunomagnetic separation enrichment process coupled with a modified real-time cellular analysis (RTCA) system (RTCA version 2) for the detection of C. difficile toxin (CDT) in stool. The limit of CDT detection by RTCA version 2 was 0.12 ng/ml. Among the consecutively collected 401 diarrheal stool specimens, 53 (13.2%) were toxin-producing C. difficile strains by quantitative toxigenic culture (qTC); bacterial loads ranged from 3.00 × 10(1) to 3.69 × 10(6) CFU/ml. The RTCA version 2 method detected CDT in 51 samples, resulting in a sensitivity of 96.2%, a specificity of 99.7%, and positive and negative predictive values of 98.1% and 99.4%, respectively. The positive step time ranged from 1.43 to 35.85 h, with <24 h for 80% of the samples. The CDT concentrations in stool samples determined by RTCA version 2 correlated with toxigenic C. difficile bacterial load (R(2) = 0.554, P = 0.00002) by qTC as well as the threshold cycle (R(2) = 0.343, P = 0.014) by real-time PCR. A statistically significant correlation between the CDT concentrations and the clinical severity of CDI was observed (P = 0.015). The sensitivity of the RTCA version 2 assay for the detection of functional toxins in stool specimens was significantly improved when the immunomagnetic separation enrichment process was incorporated. More than 80% positive results can be obtained within 24 h. The stool specimen CDT concentration derived using the RTCA version 2 assay correlates with clinical severity and may be used as a marker for monitoring the status of CDI.