A systematic review of research literature published in English between 1987 and 1997 was carried out to test the hypothesis that care planning and/or record keeping in nursing practice has no measurable effect on patient outcomes. The study was limited to research applicable to acute in-patient care and to non-information technology based recording. A search strategy was agreed with the steering group and search terms refined as the study progressed. Using the guidelines from the University of York National Health Service Centre for Reviews and Dissemination (NHS Centre for Reviews 1995), a data extraction proforma was constructed. An initial search revealed approximately 300 possible abstracts for discussion. Further analysis limited this to 43 studies, of which 30 were rejected as having too little information. Of the remaining 13 studies, none was sufficiently robust to be included in the review. The hypothesis can be neither accepted nor rejected. This has important consequences for nursing practice and management and for research. A multi-centre rigorous study is recommended.