Abstract: A multi-dimensional fragility evaluation methodology for bridge is proposed based on multiple seismic demand parameter analysis. The method incorporates uncertainties in ground motion and performance limit state (PLS) and extends the definition of fragility to multi-dimension problems. A novel probabilistic seismic demand model (PSDM) accorded with multivariate lognormal distribution is addressed to discuss the dependencies of various component responses. Considering the correlation of component PLS, the generalized multi-dimensional PLS function is established. Finally Monte Carlo simulation is performed to calculate the fragility of system. A multi-span continuous reinforced concrete girder bridge is used as an example to illustrate the approach. The samples of maximum responses are obtained through nonlinear dynamic analysis to calculate the maximum likelihood estimators of unknown parameters in PSDM, then the system fragility curve is developed and compared with individual component fragility. The result shows that multi-dimensional fragility of bridge is higher than component fragility, thus eliminating non-conservative estimation resulting from substituting the component fragility for system. The proposed method will better ensure the safety of structures, providing theoretical evidence for bridge retrofit and reliability analysis of transportation network.