Abstract Height and infant mortality are both considered health indicators of a population, yet they tend to be much more strongly correlated in high‐income, low‐mortality populations. This article shows that infant deaths are not representative of the health of survivors as it relates to height because breastfeeding practices shield them from part of the disease environment. Instead, child mortality rates, especially from food and waterborne diseases, capture the disease load that is associated with lower heights better. The period of this study is 1850–1940, with a focus on 1875–1900, as the Netherlands underwent major health and wealth transitions. Individual conscription heights from the Historical Sample of the Netherlands as well as municipal conscription statistics are used. The article takes a diachronic approach to examine how various health indicators have developed over time. The start of the upward trend in heights and the improvement of child mortality rates coincided in four Dutch regions, whereas infant mortality rates followed a different trajectory. Bivariate maps are used to identify municipalities in which infant and child mortality did not correlate. This study adds to both the understanding of heights as a health indicator and local breastfeeding practices in the nineteenth‐century Netherlands.