It is widely acknowledged that a hierarchical framework is the most appropriate way to organize a taxonomy of psychopathology and that dimensional nosological research must move beyond diagnostic comorbidity analyses in order to (a) avoid the methodological artifacts associated with diagnoses and (b) model more finely grained (and homogeneous) psychopathology dimensions. However, there has been little comprehensive work that brings these two important approaches together. This project integrates results from multiple exploratory hierarchical methods in order to facilitate the validation and interpretation of symptom-level metastructural models. Models were based on relatively comprehensive symptom-level epidemiological data from the 2000 British Office for National Statistics Survey of Psychiatric Morbidity (N = 8,405). Study 1 focuses on exploratory bifactor models of symptom-level data and the replicability of these models in a confirmatory framework. Study 2 integrates results from Study 1 with those from a bass-ackwards model. Study 3 compares results from Study 1 and Study 2 with those from corresponding models using diagnosis- and symptom count-level data.