BACKGROUND: The effectiveness of exercise for improving hand and wrist function in people with rheumatoid arthritis (RA) is uncertain. OBJECTIVES: The study aims were (1) to estimate the clinical effectiveness and cost-effectiveness of adding an optimised exercise programme for hands and upper limbs to standard care for patients with RA; and (2) to qualitatively describe the experience of participants in the trial with a particular emphasis on acceptability of the intervention, exercise behaviours and reasons for adherence/non-adherence. DESIGN: A pragmatic, multicentred, individually randomised controlled trial with an embedded qualitative study. Outcome assessors were blind to group assignment and independent of treatment delivery. SETTING: Seventeen NHS trusts in England comprising 21 rheumatology and therapy departments. PARTICIPANTS: Adults with RA who had pain and dysfunction of the hands and/or wrists and had been on stable medication for at least 3 months. Patients were excluded if they were under 18 years old, had undergone upper limb surgery/fracture in the last 6 months, were on a waiting list for upper limb surgery or were pregnant. INTERVENTIONS: Usual care or usual care plus an individualised exercise programme. Usual care consisted of joint protection education, general exercise advice and functional splinting if required. The exercise programme consisted of six sessions of strengthening and stretching exercises with a hand therapist, daily home exercises and strategies to maximise adherence. MAIN OUTCOME MEASURES: The primary outcome was the Michigan Hand Outcome Questionnaire (MHQ) overall hand function subscale score at 12 months. Secondary outcome measures included the full MHQ, pain, health-related quality of life (Short Form questionnaire-12 items), impairment (grip strength, dexterity and range of motion) and self-efficacy. European Quality of Life-5 Dimensions, medication and health-care use were collected for the health economics evaluation. Follow-up was at 4 and 12 months post randomisation. Analysis was performed on an intention-to-treat basis. RESULTS: We randomised 490 patients (244 to usual care, 246 to exercise programme). Compliance with the treatments was very good (93% of usual care participants and 75% of exercise programme participants completed treatment). Outcomes were obtained for 89% of participants at 12 months (222 for usual care, 216 for exercise programme). There was a statistically significant difference in favour of the exercise programme for the primary outcome at 4 and 12 months [mean difference 4.6 points, 95% confidence interval (CI) 2.2 to 7.0 points; and mean difference 4.4 points, 95% CI 1.6 to 7.1 points, respectively]. There were no significant differences in pain scores or adverse events. The estimated difference in mean quality-adjusted life-years (QALYs) accrued over 12 months was 0.01 greater (95% CI -0.03 to 0.05) in the exercise programme group. Imputed analysis produced incremental cost-effectiveness ratio estimates of £17,941 (0.59 probability of cost-effectiveness at willingness-to-pay threshold of £30,000 per QALY). The qualitative study found the exercise programme to be acceptable and highlighted the importance of the therapist in enabling patients to establish a routine and incorporate the exercises into their lives. CONCLUSIONS: The results of the Strengthening And stretching for Rheumatoid Arthritis of the Hand trial suggest that the addition of an exercise programme for RA hands/wrists to usual care is clinically effective and cost-effective when compared with usual care alone. No adverse effects were associated with the exercise programme. The economic analysis suggests that the intervention is likely to be cost-effective. STUDY REGISTRATION: Current Controlled Trials ISRCTN 89936343.