Soldier's heart was a medico-psychiatric condition that was first documented during the American Civil War. This condition affected British and American soldiers during the late-nineteenth and early-twentieth centuries; doctors recorded patients experiencing palpitations, breathlessness, headaches, and praecordial pain among other symptoms. While the number of cases of this disorder reached its peak in the First World War, it disappeared shortly afterwards. Based on an analysis of experimental results published in generalist and specialized medical journals as well as the correspondence between physicians and researchers that these journals maintained, this study challenges the view that soldier's heart disappeared because doctors realized that the disorder was, in fact, psychosomatic. Instead, this article shows that this notion was an unintentional by-product of the research conducted into the condition, the results of which opposed the somaticist philosophy that many of the researchers had tried to uphold.