To design and analyze chemical vapor deposition (CVD) reactors, computer models are regularly utilized. The foremost aim of this thesis research is to understand how thin film uniformity can be controlled in a CVD reactor. A complete understanding of chemical reactions that take place both in gas phase and at the deposition surface is required to predict thin film properties such as growth rate and composition precisely, however, deposition rates and surface topography can be determined by the arrival flux of reactants in a mass-transfer limited regime. In order to understand experimental thickness and roughness uniformity, a predictive model has been developed to study the fluid dynamic effect on thin film growth in a horizontal type reactor using velocity, temperature, pressure and viscosity as tunable parameters upon which velocity profiles within a CVD reactor have been evaluated using computational fluid dynamic (CFD) calculations. Through this predictive model, it is shown that fluid velocity is the major variable contributing to transverse roll cell formation compared to temperature and pressure gradients present during thin film deposition in a meso-scale CVD reactor. These results provide a physical insight regarding improved reactor operation conditions that influence uniformity.