This paper re-examines the high rates of infant mortality observed in rural areas of eastern England in the early years of civil registration. Infant mortality rates in some rural registration districts in the East Riding of Yorkshire, Cambridgeshire and Norfolk were higher than those in the mill towns of Lancashire. After describing the areas affected, this paper considers three potential explanations: environmental factors, poor-quality child care associated with the employment of women in agriculture, and the possibility that the high rates were the artefactual consequence of migrant women workers bringing their children to these areas. These explanations are then assessed using a range of evidence. In the absence of reliable cause of death data, recourse is had to three alternative approaches. The first involves the use of the exceptionally detailed tabulations of ages at death within the first year of life provided in the Registrar General's Annual Reports for the 1840s to assess whether the 'excess' infant deaths in rural areas of eastern England happened in the immediate post-natal period or later in the first year of life. Second, data on the seasonality of mortality in the 1840s are examined to see whether the zone of 'excess' infant mortality manifested a distinctive seasonal pattern. Finally, a regression approach is employed involving the addition of covariates to regression models. The conclusion is that no single factor was responsible for the 'excess' infant mortality, but a plausible account can be constructed which blends elements of all three of the potential explanations mentioned above with the specific historical context of these areas of eastern England.