One of the possibilities for improving decision processes, and the knowledge management across interacting organizations is to explore successful past experiences. Case-based reasoning (CBR) is a problem solving strategy which is based on the reuse of past solutions (cases) to address new problems. Ontologies are a means to facilitate sharing and reuse of bodies of knowledge across organizations and applications on the basis of a well-defined and precise semantics for concepts and terms. This work presents a proposal aimed at knowledge reuse, during the decision activities by means of interwoven concepts from the knowledge management, CBR and ontologies research. This blended approach presents an ontological case construction for CBR systems as theoretical and empirical support for knowledge sharing. We obtain a formal characterization of a case by means of an ontological description of particular cases, and their interrelationships with cases stored in different case-repositories. An architecture for a distributed CBR system is proposed on the basis of a multiagent setting for semantic-based access to knowledge.