n this pragmatic, multicentre, parallel-group trial, at 17 National Health Service sites across the UK we randomly assigned 490 adults with rheumatoid arthritis who had pain and dysfunction of the hands and had been on a stable drug regimen for at least 3 months, to either usual care or usual care plus a tailored strengthening and stretching hand exercise programme. Participants were randomly assigned with stratification by centre. Allocation was computer generated and unmasked to participants and therapists delivering treatment after randomisation. Outcome assessors and all investigators were masked to allocation. Physiotherapists or occupational therapists gave the treatments. The primary outcome was the Michigan Hand Outcomes Questionnaire overall hand function score at 12 months. The analysis was by intention to treat. We calculated cost per quality-adjusted life-year. This trial is registered as ISRCTN 89936343