This article aims to give guidance on the conduct and reporting of case series and audits of acupuncture, based on common problems that have arisen in the past. This type of project, and particularly the prospective case series or pre- post-intervention study, may give valuable evidence of the overall effectiveness of acupuncture--for example in different situations and in different conditions-- and provides one step in the research pathway before generating an hypothesis. The project should be designed with the aim of reducing bias as much as possible. Careful and detailed planning is essential for the project to produce worthwhile results that readers can evaluate and replicate. Ethical issues should be considered and formal approval may be necessary. The patient group should be recruited systematically and baseline data obtained. The treatment given should be systematic and decisions to change or end treatment made explicit. The outcome should be measured in ways that are known to be reliable and valid. Musculoskeletal problems can be evaluated with scales for pain and confirmed by measuring one other symptom such as stiffness or one other aspect such as bothersomeness. Global change scores also provide supporting information, and the MYMOP (Measure Yourself Medical Outcome Profile) measure is popular when patients with different conditions are included. The design of questionnaires for beliefs and attitudes is a specialised area that should not be attempted without expert help. Adverse events should also be recorded. Analysis of the data and the best way of summarising and presenting the results are also discussed.