Ontologies are widely considered as a foundational technique of the Semantic Web, in which meanings of terms are defined by formal ontologies and semantic annotations facilitate the access to Web content. This view has led to the standardization of the Web Ontology Language OWL 2. In this talk, we will reflect on the usefulness of OWL 2 ontologies for Life Sciences. We will do this by presenting a number of advantages of OWL 2 ontologies: interoperability, semantics, reasoning services. But we will also notice that it is often the case that applications only ever use terms, e.g., the classical utilization of SNOMED-CT is to use its catalogue and codes to index medical records. We will discuss a new approach reconciling ontologies and terminologies. We propose to use an OWL 2 ontology for clear semantics and reasoning, and to derive from it a lightweight terminology (or perhaps something like an OWL 2 QL ontology) for applications such as resources indexing and search. For large vocabularies, the underlying rich ontology is essential to ensure that the lightweight derived terminology is quality-controlled. Reliability is particularly important for Health Care and Life Sciences applications where safety is critical. I will illustrate this by examples, including the semantic annotation of brain MRI images (IEEE Transactions on Medical Imaging 2009), the Foundational Model of Anatomy in OWL 2 and its use for a European Portal of Health terminologies dedicated to resources indexing (AIIM 2012 under press).