This study examined the effect of three different nursing interventions in the management of women diagnosed as suffering from manic depressive psychosis. The thesis was presented in three parts. In Part One, an examination was made of the role of the nurse in caring for patients suffering from affective disorder in general. Four sub-studies were used to define the construct `routine nursing care'. These comprised: a critical appraisal of (1) the training and (2) the examination of Scottish nursing students; Scottish nurses' perceptions of their role in caring for, (3) patients with manic depresssive psychosis and (4) patients with affective disorder in general. These sub-studies suggested that psychiatric nurses were trained to offer a general, supportive, pattern of care to all patients suffering from affective disorder. No clear difference was found between the role perceptions of nurses caring for `depressed' patients in general, and those caring for manic depressive patients in particular. In Part Two, a further sub-study (5) examined a specific psychological construct, locus of control, within the context of women diagnosed as manic depressive psychosis, in remission. This study suggested that women with a history of depression-only, differed from a normative sample on one scale, with both depression-only and mania and depression subjects differing from each other, and from the normative sample on a second locus of control scale. In a final sub-study (6) a new locus of control scale was developed to measure the patient's expectations of her capacity to influence her status as a sufferer from affective disorder. Sub-studies 5 and 6 suggested the possible role played by the locus of control construct, as a mediating factor in the precipitation and maintenance of major affective disorder. In Part Three, the main (experimental) study compared the effect of three discrete nursing interventions, Routine Nursing Care, Self Evaluation and Modified Cognitive Therapy, on measures of four dependent variables characteristic of depression, and satisfaction with care and treatment. The results suggested that, despite the absence of significant between-group differences on all measures of the dependent variables, the Modified Cognitive Therapy intervention showed more clinically significant changes on three of the clinical variables, and that subjects in the MCT group become more internalised on the locus of control measure. This, suggests that the MCT group subjects' view of their capacity to control external sources of reinforcement, might have increased as a function of exposure to the training in self-management inherent in the MCT intervention. The implications of these findings for psychiatric nursing education and practice are discussed.