Explores agent-based approaches toward organizational computing models by using an organizational-learning oriented classifier system (OCS) and investigates problems in the conventional definitions/models of OL. A detailed analysis of the relationship between OCS and conventional definitions/models has revealed that (1) agent-based approaches toward organizational computing models provide detailed understanding of organizations; (2) conventional definitions/models of OL lack the concrete relationship between individual and organization.