Manipulation of the metatarsophalangeal joint and injection with steroid and local anaesthetic are widely practised in the treatment of hallux rigidus, but there is little information on the outcome. We report the results of this procedure carried out on 37 joints, with a minimum follow-up of one year (mean, 41.2 months). Patients with mild (grade-1) changes gained symptomatic relief for a median of six months and only one-third required surgery. Two-thirds of patients with moderate (grade-2) disease proceeded to open surgery. In advanced (grade-III) hallux rigidus, little symptomatic relief was obtained and all patients required operative treatment. We recommend that joints are graded before treatment and that manipulation under anaesthetic and injection be used only in early (grades I and II) hallux rigidus.