"Safety Culture" is a concept, which during recent years has been mentioned related to mari-time safety as a meta-aspect but without a comprehensive definition. This study aims at finding a workable definition of safety culture within a framework of organisational psychology. The research method was built upon a base of literature studies. The task was to perform a case study of a shipping company and one of its ships. An important element of the study was to examine how the ISM code complied with the organisational culture of the company during the implementation of the code. The case study consists of five subcases, chosen from work situations guided by the per-spective of "critical incidents" and "generative themes". The research task was systematised into five subcases. The research method implied the researcher's presence onboard during a voyage. During the voyage interviews with the crew members were carried out. Sequences of work were docu-mented by video recording and were completed by field observations and walking around observations. Observant participating in meetings and crewmembers' professional and social conversation took place. Relevant correspondence and documentation were also analysed. Daily meetings with the ship management were performed to consolidate the observations made. The company's headoffice was visited twice in order to obtain a holistic view of the interaction between shipboard management and central management functions ashore and to follow up findings from the study. The analytical work was composed of an iterative process of alternating collecting of data writing down field notes listening to recorded interviews, watching video films combined with reflection and talk within the research team. The final synthesis was the integration of the data in an explanatory framework. The conceptual or explanatory framework reflects the cultural psychological view that culture is composed of artifacts, espoused values and basic assump-tions, which together are governing the actions of the involved subjects. The analysis of the cases has also included elements of the cultural historical activity theory. Finally, the safety culture is described by eight dimensions. These dimensions express a qualitative estimation of a safety culture. They are dualistic i.e. they can be given "positive" or "negative" values with respect to an overall hypothetical concept of "good safety culture". This way to describe the safety culture is workable as language in a learning context when presenting and explaining the elements of the safety culture to the members of the organisation. Starting from the experiences, obtained from the above mentioned interventions in the onboard organisation, learning models are designed to understand and change the safety culture. As the study was performed in a shipping company, submitted to the culture of today, the question of future validity arises. Organisation has been defined as relations between organisa-tion as members. If these relations are changing to other kinds of relations by information technology, will the cultural conditions of today's learning organisation still remain? This is the question which is proposed as a task for further research work.