Abstract
It is argued that research about the Five-Factor model has much to gain from specification and analysis of middle-level categories that bridge the distance between the broad-band Five-Factor level and the adjective or item level. Two competing approaches, the AB5C-model and the NEO-PIR model, are discussed as viable but not necessarily incompatible approaches for further differentiation of the Five-Factor model. Analysis at the intermediate level of trait categorisation is considered to be essential for charting the developmental roots of the Five-Factor model and to advance cross-language comparisons of trait structure.