Previous ecological studies have suggested that early life factors are important causes of adult cardiovascular and respiratory disease, by showing geographic correlations between past infant mortality rates and present adult mortality rates. However, these studies inadequately take account of the fact that areas which were severely deprived earlier this century remain the most deprived today. Thus the ecological relation between infant and adult mortality rates could simply reflect persistence in the geographic distribution of poor socioeconomic circumstances. To explore this hypothesis further infant mortality rates for 1895-1908 for 43 counties in England and Wales were correlated with cause-specific adult mortality for 1969-73 in people aged 65-74 years, with and without adjustment for present-day social deprivation and social class. The strong simple correlations found between infant mortality in 1895-1908 and adult mortality from various causes in 1969-73 were generally much attenuated or abolished by controlling for indices of present-day socioeconomic circumstances. Our results suggest that previous studies give no strong support for any direct influence of factors acting in early life on adult coronary heart disease mortality risk. Studies which gather data about infancy, childhood, and the full course of adult life are required to clarify this issue.