Abstract: The fusion algorithm of uncorrelated measurements is extended to the case of correlated measurements.Two methods are compared.One method uses the proposed algorithm to fuse multiple correlated measurements,then uses the fused measurement in the Kalman filtering.The other uses multiple correlated measurements directly in the Kalman filtering.Equivalence of the two methods are analyzed,and the optimality of the first method is then proved.Simulation result confirms the correctness of the theoretical analysis.