Frequency Domain Decomposition(FDD)is a consolidated technique,which has a good capability of extracting modal parameters under ambient excitation.But it still needs user interaction by peak picking in singular vectors.Other than an automatic identification cannot be achieved,but gets anomalies while identify the parameters of structures with closely spaced frequencies and repeated frequencies.This research proposed an automated modal parameters identification procedure based on modal correlation between the corresponding mode shapes and FE mode shapes,namely MAC-FDD,which will conduct an automatic search respect to the target mode frequency range using only the first singular vector.