BackgroundCaries risk assessment (CRA) tools could address oral health disparities and enhance the efficiency of the oral health care system. The authors aimed to explore the feasibility and limitations of using clinical CRA tools in informing oral health care policy-making processes. MethodsThe authors used the National Health and Nutrition Examination Survey to construct 10 CRA models from a sample of clinical CRA tools identified from the literature. They used these models to estimate the proportion of publicly insured people aged 1 through 20 years categorized as at low, moderate, and high risk, and they projected their oral health care costs. ResultsThe authors found substantial variation among the selected models in assigning risk levels. The weighted average proportions (range) of people categorized as at low, moderate, and high risk were 25% (0%-66%), 14% (0%-50%), and 61% (11%-100%), respectively. Depending on the CRA model, the projected annual cost of covering this population ranged from $18 billion to $127 billion. Conclusions and Practical ImplicationsDeveloping a valid, evidence-based, accurate, and reliable population-based CRA model could address the variability among clinical CRA tools, improve estimates of dental disease burden, help design targeted oral public health programs, and enable comparative effectiveness analyses among oral health care interventions.