A comparison of two methods of criteria formulation was carried out as part of a 3-year research study looking at the effectiveness of a nursing quality assurance system (the Royal College of Nursing's Dynamic Standard Setting System). In particular, the study wished to explore whether and to what extent criteria sets derived from an expert group differed from criteria generated by local practitioner groups. One expert group and five local groups were helped to develop criteria on the topic of post-operative pain management. Results showed that the expert group underwent a much more controlled and systematic approach to the task than any of the local groups who were supported by newly trained facilitators. Despite this difference, two of the five local groups had criteria convergence scores of more than 55%, demonstrating that the local knowledge on the topic was acceptable. Whether the lower convergence of the groups was due to lack of knowledge or a combination of difficult group processes and novice facilitation is difficult to say. However, what the study did show was that local groups can generate criteria sets comparable to expert groups, given proper facilitation and support.