The aim of this study was to assess the effectiveness and healthcare costs of a practice-based osteopathy clinic for subacute spinal pain. A pragmatic RCT was carried out in a UK primary-care osteopathy clinic accepting referrals from 14 neighbouring practices. A total of 201 patients with neck or back pain of 2–12 weeks’ duration were allocated at random to usual GP care or to an additional three sessions of osteopathic spinal manipulation. The primary outcome measure was the Extended Aberdeen Spine Pain Scale (EASPS). Secondary measures included SF-12, EuroQol and Short-form McGill Pain Questionnaire. Healthcare costs were estimated from the records of referring GPs. Outcomes improved more in the osteopathy group than in the usual care group. At 2 months, this improvement was significantly greater in EASPS and SF-12 mental score. At 6 months, this difference was no longer significant for EASPS, but remained significant for SF-12 mental score. The mean healthcare costs attributed to spinal pain were significantly greater by £65 in the osteopathy group. Osteopathy also cost £22 more in mean total healthcare cost.