Conventional maturity methods estimate strength from temperature–time histories using equivalent age and apparent activation energy (Ea). However, determining Ea is time-consuming and requires extensive experimental testing. Moreover, its value depends on cement composition, mixture proportions, and curing conditions, which limits the applicability of traditional maturity-based approaches. This study proposes a new maturity-based methodology for predicting OPC concrete compressive strength development. The proposed approach eliminates both the calculation of equivalent age and the need to determine activation energy. The novelty of the work does not lie in the use of artificial neural networks (ANNs) alone. The main contribution is the integration of the maturity concept with hydration heat parameters obtained from isothermal calorimetry. Cement hydration is characterized by two calorimetric parameters: the cumulative heat released after 12 h (Q₁₂) and after 48 h (Q₄₈). These parameters describe both the rate and the extent of hydration. They are combined with curing temperature history and water-to-cement ratio in an artificial neural network model. The experimental program included seven cements produced at four cement plants. Mortars were cured under different temperature conditions and with different water-to-cement ratios. Calorimetric measurements showed that Q₁₂ and Q₄₈ were only slightly affected by changes in the water-to-cement ratio. Therefore, the same hydration descriptors could be used for a wider range of mortar compositions. The developed ANN model successfully captured the nonlinear relationship between hydration characteristics, curing conditions, and compressive strength development. Good agreement was obtained between measured and predicted strength values. The scientific contribution of this study is the development of a new maturity-based framework that replaces activation-energy-dependent formulations with calorimetrically determined hydration heat descriptors. The proposed methodology combines maturity theory, hydration heat kinetics, and artificial neural networks into a single predictive tool. This approach simplifies practical implementation and improves the prediction of strength development in cementitious materials exposed to variable curing conditions.