Like all large human-made digital artefacts, ontologies written in the Web Ontology Language OWL may contain errors, which are often detected when reasoning finds an inconsistency or generates consequences that are either unexpected or seem incomplete. In order to decide whether such a reasoning result really points to an error, the users need to be able to understand its derivation. Once they have determined that there is indeed an error in the ontology, they need help in repairing this error. After an introduction into ontology languages based on Description Logics, with an emphasis on the tractable DL EL, the course will describe classical explanation and repair methods employing axiom pinpointing. Then it will consider more sophisticated explanation approaches using proofs to explain consequences and abduction to explain non-consequences. Finally it describes repair methods that preserve more consequences than the classical ones.