GATE is currently considered in scintigraphic imaging as a powerful tool to develop, design and optimize nuclear medicine modalities. This paper describes the GATE simulation of a pixelated gamma camera which is dedicated to high resolution of small animals imaging. It consists of a CsI(Na) crystal array coupled to position sensitive photomultiplier tube. The simulation model includes photon tracking through low energy high resolution hexagonal parallel holes collimator, CsI(Na) pixelated crystal, back-compartment, and camera shielding. Simulations were compared with experimental results by some criteria such as energy spectrum, energy resolution, spatial resolution, sensitivity and count profiles obtained from line and point sources imaging. The acquired energy resolution show good agreement with measured spectra. Difference between calculated and experimental values is about 0.3% for absolute sensitivity measurement. The result of the image uniformity is more consistent after implementation of non-uniformity correction. These values were about 1.3 and 1.2% for experimental and simulation study in the central field of view, respectively. Measurements showed that the spatial resolutions differences at the head surface along the long dimensions of gamma camera for simulation and experimental differed by no more than 4%.Differences along the short axis were about 6%. The FWHMs of images of point and line sources show good consistency between experimental images and corresponding simulated ones. The difference between experimental and simulated system parameters was within 11%. Our results demonstrate the ability and flexibility of the Monte Carlo simulation for modeling pixelated gamma camera with position sensitive detector by selecting the appropriate parameters for digitizer chain and collimator position on the detector surface.