In recent years, the digitisation of historical data containing cause-of-death information has significantly increased. However, these data show considerable variations in diagnostic practices and nosology over time and place. Examining vague historical causes of death, often denoting symptoms rather than specific diseases, is a particular challenge. Infantile convulsions are an example of a common yet problematic cause of death. To improve our understanding of infantile convulsions, we propose an innovative mixed-methods, comparative approach. This study combines qualitative analyses of historical medical thinking on infantile convulsions with quantitative analyses of individual-level death records from four European cities: Amsterdam, Hermoupolis, Maastricht, and Rostock, covering different periods between 1800 and 1955. Our findings reveal that infant deaths attributed to convulsions encompass a multitude of causes from different disease categories. Significant differences emerged in the patterns of convulsions across time, age groups, and locations, even within the same country. The decline in convulsions mortality seems to be more related to the introduction of uniform registration regulations and systems, and advancements in medical knowledge than to the decline in overall infant mortality. This study’s outcome serves as a cautionary note that challenges the prevailing attitude towards convulsions and emphasises the complexity of interpreting deaths from convulsions. These were highly dependent on historical context, especially local medical culture and the variable accuracy of cause-of-death registration. These findings have implications for studies on infant mortality even when the main interest of such studies is not convulsions mortality.