We describe the process used during the international Conference on Outcome Measures in Rheumatoid Arthritis Clinical Trials (OMERACT). The objectives of the conference were (1) to broaden consensus on the minimum number of outcome measures to be included in all RA clinical trials in rheumatoid arthritis (RA); (2) to achieve consensus on criteria for (a) minimum clinically important improvement in patients with RA and (b) minimum important differences between treatment groups in RA clinical trials; and (3) to decide whether aggregate outcome measures (indices) are useful in the assessment of patients and trials. A combination of plenary sessions and structured nominal groups were employed during the conference. Simulated patient profiles and clinical trial profiles were used to generate discussion. The objective of the nominal group exercises was to capture each participant's judgments of the relative importance of each outcome measure and the degree of change required to indicate clinical improvement. Considerable discussion ensued on the content of the core set of outcome measures. An electronic interactive voting machine was used to obtain participants' views on a core set of outcome measures and methodological issues. To permit further discussion of outcome measures, the group explored the use of aggregate outcome measures (indices) in patient care and trials only in a preliminary way. A final plenary session dealt with patient perceptions of minimum important differences, a new classification of antirheumatic drugs, and a repeat of part of the preconference questionnaire. Concluding statements and future plans were developed at the conclusion of the meeting.