In this paper the approach to predict vibrations in motorcycles is presented. It can be divided mainly in two parts: prediction of engine forces using multi body dynamics (MBD) simulation and prediction of vibration response using FEA. Dynamic forces predicted at each engine mount through MBD simulation are used as input to FE analysis for vibration prediction.Single cylinder SI engine having primary balancer shaft is considered to develop this methodology. Flexibilities of important parts are considered for MBD simulation. Crankshaft ball bearing which is used in almost all two wheeler engine is modeled with 6×6 stiffness matrix. It provides coupling between radial, axial and tilting deflections of bearing and it also allows moment transfer from crankshaft to casing. This helps to predict realistic forces at each bearing and engine mounts. Distribution of primary and secondary forces at crank bearings and at different engine mounts is studied.Forces predicted from MBD simulation at each engine mounts are applied to full vehicle FE model. Forced response is predicted at each TSP and compared with test response at same point. It was found that predicted response was in well agreement with test response. Based on understanding of distribution of forces it is decided to increase mount stiffness of the support experiencing more force. This in turn helped to reduce vibration at TSP. This approach is effective to predict vibrations in early design stage to reduce development cost and time.