Pultrusion is a continuous process for manufacturing polymer composite with uniform cross-sectional profiles. In this process the pulling speed and die temperature are the main process variables that can be used to improve the chemical and mechanical properties of the pultruded polymer composite. A critical processing step in reactive polymer composites that involves thermoset resins is the curing reaction that starts from monomers/oligomers, which forms a three-dimensional cross-linked network. While empirical kinetic models for the prediction of the degree of cure are easy to handle, they are limited in terms of providing a complete understanding of the system, due to the absence of knowledge regarding the full kinetic of the functional groups. In this regard, the use of phenomenological models, based on material balances of functional groups involved in the curing reaction, is a noteworthy strategy. In this work two kinetic models were tested to simulate the pultrusion process: (i) empirical model and (ii) phenomenological model. Diffusional limitations on the cure kinetics were coupled into both models. The kinetic parameters of both models were estimated from differential scanning calorimetry (DSC) experiments of an epoxy resin derived from an unmodified liquid diglycidyl ether of Bisphenol A (DGEBA resin) in a mixture with an Anhydride Curing Agent and an Accelerator like DMP-30 (2,4,6-tris(dimethylaminomethyl) phenol). The results revealed that the kinetic models could be reasonably adjusted to the experimental curing behavior, allowing to obtain accurate values for different curing rates. The kinetic models were then implemented into the pultrusion model, by the use of the FE software, ANSYS-17.2. According to the results of ultruded thermal and curing profiles of pultruded parts, it is shown that the kinetic models are suitable for predicting the thermal and curing behavior of the pultrusion process.