Sedation which is used for intubated patients may prolong mechanical ventilation by increasing the risk of complications. The aim of the study was to illuminate the specific terminology and unrecognized contextual factors which may influence nurses' and physicians' sedation practices. The main research questions were: How do nurses and physicians describe sedation? and How does the level of nursing skill relate to the level of sedation? The hypotheses were that sedation practices are inconsistent and that experienced nurses provide a better quality of sedation than less experienced nurses. The hypotheses were supported by the study. The research strategy was case study research with triangulation of sources and methods and a multicentre multiple-case design. Four university hospitals in Copenhagen, Denmark, and 14 cases were included in the study. The findings were based on secondary analysis of observation, interviews and chart review. The theoretical framework for the study was the problem-solving model, in which sedation was assumed to be provided according to indication (clinical problem), intervention (clinical decision) and expected outcome (clinical end-point). Indications could be patient-related, ventilator-related, or patient-ventilator related. Interventions could be related to the choice of agent, dose or administration method and the outcome was the level of sedation. Sedative therapy was prescribed by physicians and administered by nurses. The four sites in the study did not use guidelines for sedation and did not use sedation level assessment tools. The study shows that when the terminology is unclear, the indications, interventions and outcomes become unclear.