The author reviews supporting evidence for the existence of two personality prototypes that correspond to two DSM-III-R personality "clusters." These hypothetical personality prototypes are characterized by extreme variations along several interrelated personality trait dimensions and by differences in two biological measures. The possibility that individual differences in central serotonin and/or catecholamine activity may underlie some of the features associated with these prototypes is discussed in light of current data. The usefulness of this conceptualization as a supplemental axis of personality classification is suggested.