When, in 1955, Court Brown and Abbatt, and van Swaay independently reported that patients with ankylosing spondylitis who had been treated by X-ray therapy developed an unusually high incidence of leukaemia, it was evident that they might form a particularly suitable group for studying the long-term effects of irradiation. A large number of patients had been treated over the preceding 20 years, the dose of radiation varied over a wide range, the patients were kept under prolonged medical observation, the general mortality rate among them was not high, and there were not thought to be any complications of the disease that would mask the diagnosis of leukaemia or other cancers. The principal disadvantage was that they were likely to be given drugs, some of which might conceivably have an independent carcinogenic effect.