The aims of this thesis were to explore nurses' experiences of stress in the ICU (I), to analyze experiences of moral concerns in intensive care nursing from the perspective of relational ethics (II), to describe the synthesis of the concept of moral stress and to identify preconditions for moral stress (III) and to analyse and describe lived experiences of support in situations characterized by critical care situations and moral stress in intensive care (IV).The design was exploratory and descriptive. Material in studies I, II and IV consisted of interviews with intensive care nurses (10 head nurses and 26 staff nurses) employed in general, thoracic and neonatal intensive care units in five hospitals located in different parts of Sweden. The material in study III data from two studies of professional issues in nursing were used for the analysis: one concerned psychiatric nursing and the other was the previously referred study I.In study I qualitative content analysis and descriptive interpretation was used in the analysis. The main theme 'stress induced by dissonant imperatives' formulated in the analysis. Dissonant imperatives are composed of the four sub-themes: 1) controlled by the working situation - needing to be in control, 2) constrained by prioritisation - wanting to do more, 3) lacking authority to act - knowing that something should be done, and 4) professional distance - interpersonal involvement. In study II qualitative content analysis and descriptive interpretation were used in the analysis. A main theme was formulated, 'caring about-caring for: tensions between moral obligations and work responsibilities in intensive care nursing'. Five sub-themes were formulated 1) believing in a good death, 2) knowing the course of events, 3) feelings of distress, 4) reasoning about the physicians and 5) expressing moral awareness. In the study III a hypothetical-deductive method was used. The findings indicate that moral stress is independent of context-given specific pre-conditions: 1) nurses are morally sensitive to the patient's vulnerability, 2) nurses experience external factors preventing them from doing the best for the patient, and 3) nurses feel that they have no control over the situation. In the study IV an interpretive method was used. The first level of analysis of data identified contextual factors, such as type and purpose of support and working conditions. Thereafter five tentative interpretations were revealed: 1) receiving organised support is a matter of self-determination, 2) whether to participate or to be off duty is experienced mutually as exclusive, 3) dealing with moral stress is experienced as a private matter, 4) colleagues managing moral stress serve as models in stress support, and 5) not being able to deal with moral stress urges one to seek outside support. A comparison of these interpretations identified three major themes: availability, accessibility and receptivity of support. The main interpretation of data was: "lived experience of moral stress support involves an interconnectedness between structural and existential factors".A comprehensive understanding was formulated using the four studies (I, II, III and IV). Moral stress was found to be influential on the caring competence. Conflicts between different competences were found leading to a shift in focus away from the patients leading to a possible decrease in the caring competence. Moreover, the subtle resistance among nurses toward participation in organized moral stress support may obstruct the development of nurses' caring competence. Accordingly, imbalance, due to moral stress, between different competences hinders the development of collectively shared caring competence.