In this paper two models present in the computational tool Monte Carlo Risk Assessment (MCRA) were compared for assessing the usual intake of lead in five countries. For this, we used national food consumption data organised according to the format of the European Food Safety Authority (EFSA) Comprehensive database and a single lead concentration database in which analysed commodities were organised according to EFSA's Standard Sampling Description (SSD) system. This meant that both input data were coded according to the hierarchical FoodEx1 classification system. We demonstrate that the naïve Observed Individual Means model resulted in more conservative estimates of the exposure in the right tail of the exposure distribution compared to a refined usual intake model, the LogisticNormal–Normal model. With MCRA, the usual intake could be estimated with both models using food consumption and concentration data that were coded according to the hierarchical FoodEx1 classification system demonstrating that this tool can be used in EFSA's data environment. Additionally, the computational tool has functionalities 1) to check the input data quality by presenting detailed information about these data around a specified percentile of exposure and 2) to decide whether the use of a more refined usual intake model is appropriate.