A computational code is developed to help identify metal absorption lines in high resolution QSO spectra,especially in the Lyα forest.The input to the code includes a list of line central wavelengths,column densities and Doppler widths.The code then searches for candidate metal absorption systems and assesses the probability that each system could be real.The framework of the strategy we employ is described in detail and we discuss how to estimate the errors in line profile fitting that are essential to identification.A series of artificial spectra is constructed to calibrate the performance of the code.Due to the effects of blending and noise on Voigt profile fitting,the completeness of the identification depends on the column density of absorbers.For intermediate and strong artificial metal absorbers,more than 90％ could be confirmed by the code.The results of applying the code to the real spectra of QSOs HS0757+5218 and Q0100+1300 are also presented.