In this study, a novel chemometric algorithm is presented to facilitate the comparison of relevant chemical components from different herbal samples. This so-called multicomponent spectral correlative chromatography (MSCC) is developed to detect and decide whether two chromatographic clusters are correlated spectrally with each other. The target chromatographic cluster is first partitioned from one herbal spectrochromatogram obtained by hyphenated chromatography. Then, a projection operator is constructed with the principal spectral features extracted from the target to judge the presence or absence of a spectral correlative chromatographic cluster within another herbal spectrochromatogram. For this judgment, congruence coefficient between the original spectral vector and its projected residual is proposed to eliminate the influences from background and noises, especially heteroscedastic noises in the original data. The performance of the MSCC algorithm is demonstrated on both simulated data and real data, and its advantages and disadvantages are also discussed in some detail.