This work presents a comparison of the non-equilibrium energy transfer and dissociation of nitrogen molecules using two different approaches: the Direct Molecular Simulation (DMS) method and the State-to-State (StS) method. The two methods are used to study the thermochemical relaxation in a zero-dimensional isochoric and isothermal reactor. Both methods make use of the same potential energy surface (PES) for the N2(1Σ+g) − N(4Su) system taken from the NASA Ames quantum chemistry database. The DMS method is a stochastic method, which relies on sampling scattering calculations to directly determine the non-equilibrium evolution of a gas. In contrast, the StS model is a deterministic approach which relies on a complete kinetic database to directly solve the master equation for each internal energy state. The analysis reveals several differences between the two approaches. First, because of the reliance on sampling in the DMS method, accurately predicting the population of high energy states is difficult and subject to statistical error. Moreover, the DMS method does not map post-collision energies back to discrete states, meaning that distribution of states across energy is much smoother with this method. Despite the differences observed in the microscopic distribution, the macroscopic properties predicted by both methods agree well during both energy transfer and dissociation.