Over a thirty-year period in Scunthorpe, during which there were two changes in the sodium content of the water supply, there were significantly more sudden deaths in infancy during the decade when sodium levels were highest. This is compatible with the hypothesis that hypernatraemia is a factor leading to cot death. Changes in the incidence of breastfeeding over the same period may have contributed to, but are unlikely to account for, the result because the pattern was different in a neighbouring district where there was no change in water quality.