The present invention provides a method and an apparatus for analyzing mass analysis data for selecting a component similar to a target component quickly and accurately, based on MS.sup.n analysis for each unknown component in the sample. First, MS.sup.n analysis is performed for each of the components in the mixed sample, and based on the obtained spectral data (measured data) and the spectral data of the target component (reference data), predetermined parameters are extracted (Step S11 to S13). Next, by multivariable analysis of the parameters, the similarity between the target component and each of the components in the mixed sample is evaluated (Step S14). Finally, based on the similarity value, components similar to the target component are selected (Step S16).