Objectives: In order to distinguish similar memories, it is experimentally confirmed that the hippocampus forms distinct representations of them. The ability of the brain to disambiguate memories is known as pattern separation. It has been proposed that dentate gyrus (DG) accomplishes this task, specifically through its principal cells, called granule cells (GCs). In this project we investigate the role of GC dendrites in pattern separation by modifying their biophysical and morphological characteristics. Methods & Results: We have implemented a morphologically simple, yet biologically relevant, computational model of the DG that implements pattern separation. The network consists of four well-studied neuronal types: granule, mossy, basket, and HIPP cells. The GC model consists of an integrate-and-fire somatic compartment connected to a variable numbers of active dendritic compartments. For simplicity reasons, without sacrificing detail, we used point neurons to simulate the remaining neuronal types. GCs major input from the Entorhinal Cortex (EC) is simulated as independent poisson spike trains at realistic firing frequencies. The output of the network corresponds to the spiking activity of GCs and is estimated on two highly overlapping input patterns. Pattern separation is accomplished when the similarity between these input patterns is greater than the similarity between the respective output patterns, as assessed by the Hamming Distance (HD) metric. Preliminary results show that there is a positive correlation between the separation efficiency and the number of GC dendrites. Conclusions: Our preliminary results suggest that dendrites of GC cells facilitate the pattern separation capabilities of the DG.