AIMS: The aims of the study were to explore expert opinion on the distinction between "research" and "audit", and to determine the need for review by a National Health Service (NHS) Research Ethics Committee (REC). BACKGROUND: Under current guidelines only "research" projects within the NHS require REC approval. Concerns have been expressed over difficulties in distinguishing between research and other types of project, and no existing guidelines appear to have been validated. The implications of this confusion include unnecessary REC applications, and crucially, the potential for ethically unsound projects to escape review. METHODS: A three-stage Delphi method was chosen to explore expert opinion and develop consensus. Stage 1 comprised ten semi-structured interviews gathering opinion on distinguishing between types of project and how to determine need for ethical review. Stages 2 and 3 were questionnaires, asking 24 "experts" to rate levels of ethical concern and types of project for a series of questions. Anonymised responses from stage 2 were fed back in stage 3. The final responses were analysed for consensus. RESULTS: Of 46 questions, consensus was achieved for 14 (30.4%) for level of ethical concern and for 15 (32.6%) for type of project. CONCLUSIONS: Several ideas proved discriminatory for classifying the type of project and assessing level of ethical concern, and they can be used to develop an algorithm to determine need for ethical review. There was little relationship between assessment of the level of ethical concern and classification of the project. There was inconsistency in defining and classifying studies as something other than "research" or "audit".