Current dimensional taxonomies of personality disorder show a stronger empirical grounding than categories, but may lack the necessary level of detail to make accurate predictions and case formulations. We need to further develop the lower levels of the hierarchy until reaching the building blocks of personality pathology. The Dimensional Assessment of Personality Pathology (DAPP-BQ) is well-suited to this purpose due to its multilayered structure and its agreement with the official dimensional classifications. We disaggregated the 18 DAPP-BQ mid-level facets through exploratory and confirmatory factor analysis in a sample of 3,233 community subjects and outpatients. We obtained a set of 72 clinically relevant, narrower subfacets which were reliable, well-fitted to the data, and invariant between clinical and community subjects and between the sexes. This third level of abstraction increases by 4.7% the capacity to predict DSM categorical personality disorders, gives a particular advantage in capturing dependent, histrionic, paranoid, obsessive, and schizoid features, and can provide the detailed information that clinical decisions demand.