The present paper deals with the development and evaluation of a new semi-empirical, pseudo-multi-zone model capable of estimating NOx emissions for various types of diesel engines and also different engine configurations. The specific model is physically based due to the use of the first thermodynamic law and the consideration of combustion chemistry and dissociation of the combustion products during the closed part of the engine cycle. The model estimates the fuel burning rate through Heat Release Rate Analysis of the measured cylinder pressure which is then coupled to a simplified multi-zone approach, assuming that each element of fuel burns individually at controlled conditions having from this point on its own history inside the combustion chamber. From this procedure, a simplified multi-zone semi-empirical model is developed, that accounts for the temperature distribution inside the combustion chamber and its evolution during an engine operating cycle. On the other hand, O₂ availability is accounted for though the mean equivalence ratio value and the instantaneous gas composition (i.e., EGR). Nevertheless, to qualitatively improve calculated NOx emissions, an empirical correlation is used in the simulation for the estimation of equivalence ratio inside each fuel package. In this empirical formula is used the mean equivalence ratio and a theoretical constant which represents the equivalence ratio that results to a peak value for NOx emissions. For the quantitative improvement of calculated results, a constant scaling factor was finally used. The major inputs for the model are the measured pressure diagram, the basic engine geometry, engine speed, fuel consumption and intake air characteristics. The model was validated on two different diesel engine types: a DI medium-duty truck engine and a small passenger car engine with pilot fuel injection. For the first engine, the ability of the model to estimate NOx at all load points foreseen in the European Stationary Cycle (ESC) was evaluated. For the second engine, model"s predictive ability is examined as far as the effect of common rail injection pressure and Exhaust Gas Recirculation rate (EGR) is concerned at various speeds and loads. As revealed, the obtained results indicate a good agreement against measured NOx tailpipe values and thus the proposed model appears to be a promising tool for the development of a fast and reliable NOx model that can be coupled to simple performance models. Furthermore, the proposed model can be used for the prediction of NOx emissions in cases where cylinder pressure measurement is available. The last is now becoming more and more standard practice especially in slow speed large marine engine and engines used for power generation. NOx emissions are important and in most cases they can be controlled if their formation mechanism is properly estimated.