In the next-generation risk assessment (NGRA) of skin sensitization, estimating the point of departure (PoD) is crucial. The murine local lymph node assay (LLNA) has been considered the 'gold standard' for evaluating the skin sensitizing potential of chemicals, with the LLNA EC3 values serving as the PoD for dermal quantitative risk assessment (QRA). This study presents artificial neural network (ANN) models that predict EC3 values, enhanced by integrating the Amino Acid Derivative Reactivity Assay (ADRA) to expand the applicability domain. Initially, descriptors derived from ADRA, based on both molar and gravimetric concentrations, showed significant correlations with LLNA EC3 values. We then constructed prediction models using ANN analysis, incorporating parameters from GL497-adopted methods. These models exhibited a strong correlation with LLNA EC3 values. The predicted EC3 values for molar and gravimetric concentrations correlated well with each other and with previous values from an ANN model using DPRA instead of ADRA. Additionally, the prediction accuracy of ANN models combined with "2 out of 3″ negative judgment for GHS classification was comparable to that of ITSv1/v2. Ultimately, this enables QRA for a broader range of substances using predictive EC3 values as PoDs without animal testing, paving the way for more effective risk assessments.