The forward and inverse problems in damage quantitative identification for rolling bearing are studied. Multibody dynamics of bearing are developed according to the vibration transmission path combining with contact mechanism of interface and tribological system. The model incorporates the contact elastic deflections and forces, contact slip, contact separations, and localized faults. Dents and pits on inner race and outer race and balls were modeled through surface profile changes. Bearing load zones under various radial loads and clearances are simulated. By the aid of normal signal, model coefficients are modified and have a better performance in predicting damage. Then sensitive features matrix using Kurtosis-Laplace wavelet is accurately constructed. Through translating the inverse problem into geometric similarity matching with grey incidence degree, measured indicators are used in damage detection process and the damage size can be determined. Practical experimental data verifies the validity and robustness of the quantitative identification method, which is feasible for practical application to diagnosis defects in bearing system.