There has been a marked increase in the reported mortality from motor neuron disease (MND) but not multiple sclerosis (MS) in England and Wales and in a number of other countries. A comparison has been made of the mortality from MND and from MS for two time periods in Australia, New Zealand and South Africa. An increase in MND mortality occurred in Australia and New Zealand between 1968-77 and 1978-87, greater than that which occurred in England and Wales, but there was no increase in MS mortality. Among the white population of South Africa, the MND mortality was half of that in England and Wales, Australia and New Zealand in both time periods. Both MND and MS mortality is higher in the English-speaking than in the Afrikaans-speaking white South African-born. The marked increase in MND mortality which has now been reported from many countries, is good evidence that an environmental factor is important in causing this disease. The large differences in MND mortality in different populations may be important clues to the environment factors causing the disease. (7 Neurol Neurosurg Psychiatry 1993;56:633-637) study, 1978-87 and also for the two subsequent years, 1988-89. A total of 94 4% of the MND and 89% of the MS death certificates were traced. Previous studies on MS prevalence had shown that MS was more common among the English-speaking than among the Afrikaans-speaking white South African-born and perhaps this also occurred in MND.10To find out how frequently MND was noted on the death certificate, after it had been diagnosed during life, we obtained lists of patients diagnosed as having MND from the teaching hospitals of Groote Schuur and Belville, at the Cape, and at the General Hospital, Johannesburg. We ascertained from the Population Registrar which of these patients had died and how the death had been reported on the death certificate. The death certificates of 126 white patients diagnosed as having MND at three teaching hospitals in South Africa were traced. Among them, MND was reported as the primary cause of death in 81 or 64-3%, and as a contributory cause of death in 16 or 12-7%, compared with 68-3% and 113% in England and Wales." Previous studies have shown that MS is reported on the death certificate as frequently, when diagnosed during life, in South Africa'0 as in England and Wales."Results were expressed as standardised mortality ratios (SMR) with rates from England and Wales taken as the standard. Confidence limits for SMRs were based on the exact method using the Poisson distribution. Statistical significance of an SMR was inferred if the relevant confidence interval (95% for p < 0-05; 99% for p < 0.01) did not include the null value of 100. Confidence intervals for the ratio of two SMRs in different groups was made using the binomial dis-633 on 10 May 2018 by guest. Protected by copyright.