We have previously reported on the causes of death among 2,068 patients treated with X irradiation for metropathia haemorrhagica at three Scottish radiotherapy centres between 1940 and 1960 (Doll and Smith, 1968). This cohort of women has now been followed up for a further seven years. 500 (24 per cent) women have now died, 78 (3-8 per cent) have emigrated and 25 (1-2 per cent) could not be traced. The numbers of deaths from different causes have been compared with the numbers expected in a population of similar age and sex exposed to the Scottish national mortality rates over the same period. An excess of deaths from leukaemia (seven observed, 2-3 expected) and of cancers of the heavily irradiated sites (59 observed, 40-1 expected) continues to be observed five or more years after treatment. There is no indication of any change in the excess death rate, due to cancers of sites in the radiation field, with time since treatment up to at least 20 years after the radiation exposure. Over the same period the number of deaths from cancer of the breast was below expectation (ten observed, 22-3 expected) and no increased mortality from coronary disease was seen (102 observed, 100-9 expected). The mean dose of radiation to the bone marrow has been determined for each woman ant it is estimated that the excess rate of leukaemia in the first 20 years after treatment is about 1-1 per million women per year per rad. This figure is in accord with the estimates derived from the survivors of the atomic bomb explosions in Hiroshima and Nagasaki and among patients with ankylosing spondylitis treated with X irradiation. However, the finding of no excess risk of leukemia among women treated with irradiation for cancer of the cervix (Hutchison, 1968) suggests that the simple assumption of a linear dose-response relationship for leukaemia is incorrect, at least when high doses of radiation are delivered to a small volume of marrow.