The present study employed multivariate genetic item-level analyses to examine the ontology and the genetic and environmental etiology of the Big Five personality dimensions, as measured by the NEO Five Factor Inventory (NEO-FFI) [Costa and McCrae, Revised NEO personality inventory (NEO PI-R) and NEO five-factor inventory (NEO-FFI) professional manual, 1992; Hoekstra et al., NEO personality questionnaires NEO-PI-R, NEO-FFI: manual, 1996]. Common and independent pathway model comparison was used to test whether the five personality dimensions fully mediate the genetic and environmental effects on the items, as would be expected under the realist interpretation of the Big Five. In addition, the dimensionalities of the latent genetic and environmental structures were examined. Item scores of a population-based sample of 7,900 adult twins (including 2,805 complete twin pairs; 1,528 MZ and 1,277 DZ) on the Dutch version of the NEO-FFI were analyzed. Although both the genetic and the environmental covariance components display a 5-factor structure, applications of common and independent pathway modeling showed that they do not comply with the collinearity constraints entailed in the common pathway model. Implications for the substantive interpretation of the Big Five are discussed.