Objectives: Abundance of S. mutans in saliva is an important microbiological parameter for assessing caries risk. A novel monoclonal antibody-based immunoassay (Saliva-Check mutans) was developed and outperformed the conventional culture-based assay (Dentocult SM) in enumerating S. mutans in saliva. This study aimed to evaluate and compare their validity in predicting early childhood caries (ECC). Methods: With ethical approval and parental written consent, 190 preschool children aged 3-4 years were recruited. The abundance of S. mutans in their stimulated saliva samples was classified by the culture-based assay into four levels (0-3), while the immunoassay categorized the samples into two groups with high and low S. mutans levels. Children's tooth status was assessed at baseline and after 12 months. Caries increment (△dmft) was recorded. The validity of both assays in predicting new caries (△dmft>0) was evaluated and benchmarked against that of "past caries", which is regarded as the strongest indicator for future caries. Results: The mean caries increment was higher in children with "past caries" than in those caries-free at baseline (1.46 vs. 0.25; p<0.001). Compared with children classified by the immunoassay as harboring low level of S. mutans, children with high level of S. mutans developed more new caries (1.66 vs. 0.27; p<0.001). In children whose S. mutans level was scored as 0 to 3 by the culture-based assay, the mean caries increments were 0.31, 0.42, 1.00, and 1.56, respectively; significant differences existed only between some groups (score 0 and 2/3; score 1 and 3; all p<0.05). The sensitivity/specificity of the immunoassay, culture-bases assay, and "past caries" in predicting ECC were 0.70/0.92, 0.70/0.74, and 0.70/0.83, respectively. Conclusions: The novel immunoassay appeared to be more specific than the culture-based assay and "past caries" in predicting ECC and may potentially enhance multifactorial caries prediction (Supported by HHSRF#07080741 and CRCG#200907176096).