Background Reports of methotrexate for chronic urticaria are anecdotal. Objectives To assess the effectiveness of methotrexate in steroid-dependent chronic urticaria, its impact on steroid reduction and any differences in response between patients with and without functional autoantibodies. Methods A retrospective case-note review of 16 patients with steroid-dependent chronic urticaria treated with methotrexate was carried out. Ten patients had chronic ordinary/spontaneous urticaria (CU), including three with associated delayed-pressure urticaria; four patients had normocomplementaemic urticarial vasculitis (UV); and two patients had idiopathic angio-oedema without weals. Median disease duration before methotrexate was 48.5 months (range 12-164). All were unresponsive to antihistamines and second-line agents, except prednisolone. Eleven were assessed for autoimmune urticaria with the basophil histamine release assay (n = 5), autologous serum skin test (n = 5) or both (n = 1). Response to methotrexate was scored: no benefit; some benefit (fewer weals and symptomatic improvement but no steroid reduction); considerable benefit (improvement with steroid reduction); or clear (no symptoms, off steroids but on antihistamines). Results Twelve of 16 patients (eight CU, three UV, one idiopathic angio-oedema) responded. Three showed some benefit, seven considerable benefit and two cleared. Four of eight responders and three out of three nonresponders showed evidence of functional autoantibodies. The dose to achieve a steroid-sparing effect was 10-15 mg weekly (cumulative dose range 15-600 mg, median 135 mg). Methotrexate was well tolerated. Conclusions Methotrexate may be a useful treatment for steroid-dependent chronic urticaria. Functional autoantibodies do not correlate with response. The beneficial effects of methotrexate may be anti-inflammatory and immunosuppressive. It may therefore benefit chronic urticaria independently of the pathogenic mechanism, whether autoimmune or not.