Computer-based simulations of human hip joints generally include investigating contacts happening among soft or hard tissues during hip movement. In many cases, hip movement is approximated as rotation about an estimated hip center. In this paper, we investigate the effect of different methods used for estimating hip joint center of rotation on the results acquired from hip simulation. For this reason, we use three dimensional models of hip tissues reconstructed from MRI datasets of 10 subjects, and estimate their center of rotation by applying five different methods (including both predictive and functional approaches). Then, we calculate the amount of angular and radial penetrations that happen among three dimensional meshes of cartilages, labrum, and femur bone, when hip is rotating about different estimated centers of rotation. The results indicate that hip simulation can be highly affected by the method used for estimating hip center of rotation. However, under some conditions (e.g. when Adduction or External Rotation are considered) we can expect to have a more robust simulation. In addition, it was observed that applying some methods (e.g. the predictive approach based on acetabulum) may result in less robust simulation, comparing to the other methods.