This study investigates differentials in the decline of cause-specific infant mortality by marital status of the mother in Stockholm (1878 –1925) and factors contributing to the explanation of these differentials using computerized records of individual entries from the Roteman Archives. Included in the analysis were 120,094 children less than 1 year of age who lived in Södermalm during this period. Cause-specific mortality rates were calculated for three time periods. Cox's regression analysis was used to study the relationship between overall and cause-specific risk of infant death and of being born in and out of wedlock in relation to a set of variables. Infant mortality rates and mortality risks were higher among children born out of rather than in wedlock. The most pronounced differentials in cause-specific mortality rates between these groups of children were seen in cases of diarrhea. The socioeconomic status of the household head and number of children in the household were statistically significant with infant mortality, but explain only part of the excess mortality risk of children born out of wedlock. In Stockholm at the turn of the 19th century being born out of wedlock was strongly associated with poor health outcomes, particularly in diarrheal diseases, pneumonia/bronchitis, and immaturity/congenital causes.