The feature scale model for selective chemical vapor deposition process was proposed, and the simulation was performed to study the selectivity and the uniformity of deposited thin film using the Monte Carlo method and string algorithm, Elementary steps such as direct deposition, re-emission, and surface diffusion mere included in the model. The effect of model parameters such as sticking coefficient of the chemical precursor, aspect ratio, surface diffusion coefficient, and lateral wall slope angle was examined in detail. The thickness uniformity of selectively deposited thin film inside the trench was better in the deposition on the microstructure with higher aspect ratio and with the chemical precursor with lower sticking coefficient. It was revealed that the surface diffusion of adsorbed reactant led to the selectivity loss. With the decrease of the lateral wall slope angle, more uniform deposition profile was obtained except near the sidewall of the trench. It was found that low sticking coefficient with optimum surface diffusion coefficient gave the most uniform and selective him deposition.