A 3-year placebo-controlled trial of sulphasalazine (SASP) in 89 patients with established AS (including radiological sacroiliitis) showed a reduced frequency of peripheral arthritis in the treated group but did not show any definite benefit in the maintenance of spinal mobility. Adverse effects causing treatment withdrawal occurred in five placebo-treated patients and eight SASP-treated patients, but 22 patients preferred to stop taking daily medication of unproven benefit for the full 3 years. The natural history of established AS suggested two groups of patients: the majority with principally spinal symptoms and infrequent peripheral arthritis or iritis, and a minority who tend to have recurrent extra-spinal problems.