Objective To estimate the clinical effectiveness and cost-effectiveness of three methods of ankle support compared with double layer tubular compression bandage. Design A randomised controlled trial, designed to reflect practice in UK hospital emergency departments. Setting Eight emergency departments in England. Participants Aged 16 or over with acute severe ankle sprain, unable to weight bear, no fracture. Interventions 584 participants were randomised to one of four treatment arms: tubular bandage, below knee cast, Aircast ankle brace or Bledsoe boot, all applied 2-3 days after presentation to allow swelling to resolve. Main outcome measures Response to treatment was assessed using the Foot and Ankle Outcome Score and generic measures (Functional Limitations Profile, SF-12 and EQ-5D). Results When adjusted for age, sex and baseline scores, the below knee cast offered a small but statistically significant benefit at 4 weeks in terms of pain (FAOS pain difference 5.1; 95% CI 0.4-9.8), foot- and ankle-related quality of life (QoL) (FAOS QoL difference 5.9; 95% CI 0.1-11.8) and the physical component of the SF-12 (SF-12 score difference 2.2; 95% CI 0.0-4.4). Neither the Aircast brace nor the Bledsoe boot was statistically or clinically better. At 12 weeks the below knee cast was significantly better than tubular bandage in terms of pain (FAOS pain difference 5.1; 95% CI 0.3-10.0), activities of daily living (FAOS ADL difference 3.5; 95% CI 0.4-6.6), sports (FAOS sports difference 8.7; 95% CI 1.6-15.7) and QoL (FAOS QoL difference 8.7; 95% CI 2.4-15.0), and the Aircast brace was better only in terms of ankle-related QoL and mental health. The Bledsoe boot conferred no significant advantage over tubular bandage. By 9 months there were no significant differences. Based on mean direct health-care costs per participant, the Bledsoe boot was the most expensive (215 pounds) and tubular bandage the least so (1 pound 44 pence). Inclusion of indirect costs (sick leave) raised overall costs substantially and removed any significant differences between the therapies. Cost-utility analysis demonstrated that the Aircast brace [301 pounds per quality-adjusted life-year (QALY)] and below knee cast (339 pounds per QALY) were more cost-effective than the Bledsoe boot (2116 pounds per QALY). However, inclusion of indirect costs produced different rank orders, depending on the assumptions made, and results should be treated with caution. Conclusions The below knee cast and the Aircast brace offered cost-effective alternatives to tubular bandage for acute severe ankle sprain, the former having the advantage in terms of overall recovery at 3 months. As there were no differences in long-term outcome, practitioners should consider likely compliance and acceptability to patients when choosing a brace.