QUESTIONS: Do patients with a severe ankle injury (baseline ankle function score < or = 40) do worse in the short-term than patients with a mild injury (score > 40)? Does physiotherapy intervention have more effect on patients with a severe injury compared with a mild injury in the short- or long-term? Is self-reported recovery related to ankle function score over time? DESIGN: Subgroup analysis of a randomised trial. PARTICIPANTS: 102 adults with an acute lateral ankle sprain. INTERVENTION: The experimental group received physiotherapist-supervised exercises in addition to conventional intervention. OUTCOME MEASURES: Outcomes were self-reported recovery, pain, and instability all measured on a 10-point visual analogue scale, and incidence of re-sprain. Measurements were collected at baseline, 4 and 8 weeks, 3 and 12 months. RESULTS: Participants with a severe injury did worse in 3 out of 7 outcomes than those with a mild injury at 4 weeks but not at 8 weeks. There was no difference in effect of physiotherapy intervention in those with a severe injury compared with a mild injury, at 8 weeks or 12 months. However, there was an effect of physiotherapy intervention in those with a severe injury in 3 out of 7 outcomes at 8 weeks. Self-reported recovery was related to ankle function score at all points in time (r = 0.48 to 0.79). CONCLUSION: The results of this study only partially support the recommendations regarding the use of the ankle function score in the 'Acute Ankle Injury' guideline of the Royal Dutch Society of Physiotherapists.