An objective correlation analysis based on the concept of singular-value decomposition of a matrix is proposed here for the field of metallic traces in the atmospheric aerosol. On the basis of an atmospheric sampling at Cap Ferrat (southeastern coast of France), the method is applied to the airborne concentrations of Al, Pb, Cd, Cu, and Zn.The data matrix is decomposed in two series of singular vectors. These vectors are orthogonal and classified in decreasing importance, according to the percentage of the total variance that they explain. Such a method is easy to apply and, if compared with a standard correlation analysis, it exhibits such advantages as (i) atypical points can be objectively discarded in order to improve the description of the general characteristics of the data set; (ii) all the elements are simultaneously taken into account by the analysis, which permits the enhancement of the features of the data set involving one or several metals; (iii) the importance of these independent features in the variability of the data set is measured.