In this article, we analyse infant mortality in Sundsvall 1860–1892. The focus is on the pattern and development of causes of death with separate analyses of neonatal and post-neonatal mortality. Furthermore, we discuss the development of infant mortality in relation to possible determinants in the historical context in Sundsvall. The results show substantial differences between neonatal and post-neonatal mortality when it comes to causes of death as well as their seasonal pattern. For deaths in the first 28 days, a large proportion of the deaths were diagnosed as unknown disease or given vague and symptom-descriptive diagnoses. For post-neonatal mortality on the other hand, the dominant cause of death categories were water-and food-borne infections and air-borne infections. Water-and food-borne diseases had a very strong seasonal pattern with the peak in late summer — July and August. There is no indication that sanitary improvements in the 1880's led to fewer cases of diarrhoea. Mortality from air-borne diseases on the other hand was lowest during summer, instead peaking in the winter months.