A system and method are disclosed for automatically classifying plaque lesions. A plaque classification application applies a plaque classification criterion to at least one graphical image, comprising a map of spectrally-analyzed characterized tissue of a vessel cross-section, to render an overall plaque classification for the slice or set of slices, covering a 3D volume. The plaque classification is based upon the amount and location of each characterized tissue type (e.g., necrotic core--NC). In an exemplary embodiment the set of potential plaque classifications, not to be confused with characterized tissue types--from which the plaque classifications are derived--include, for example: adaptive intimal thickening (AIT), pathological intimal thickening (PIT), fibroatheroma (FA), thin-cap fibroatheroma (TCFA), and fibro-calcific (FC).