The Phenoscape Knowledgebase (KB) is an ontology-driven database that combines existing phenotype annotations from model organism databases with new phenotype annotations from the evolutionary literature. Phenoscape curators have created phenotype annotations for more than 5,000 species and higher taxa, by defining computable phenotype concepts for more than 20,000 character states from over 160 published phylogenetic studies. These phenotype concepts are in the form of Entity-Quality (EQ) compositions which incorporate terms from the Uberon anatomy ontology, the Biospatial Ontology (BSPO), and the Phenotype and Trait Ontology (PATO). Taxonomic concepts are drawn from the Vertebrate Taxonomy Ontology (VTO). This knowledge of comparative biodiversity is linked to potentially relevant developmental genetic mechanisms by importing associations of genes to phenotypic effects and gene expression locations from zebrafish (ZFIN), mouse (MGI), Xenopus (Xenbase), and human (Human Phenotype Ontology project). Thus far, the Phenoscape KB has been used to identify candidate genes for evolutionary phenotypes, to match profiles of ancestral evolutionary variation with gene phenotype profiles, and to combine data across many evolutionary studies by inferring indirectly asserted values within synthetic supermatrices. Here we describe the software architecture of the Phenoscape KB, including data ingestion, integration of OWL reasoning, web service interface, and application features.