After travelling in subSaharan Africa, an area known for sporadic cases of Marburg virus infection, a young Swedish man presented with a classical picture of severe viral haemorrhagic fever complicated by disseminated intravascular coagulation and septicaemia. Serum samples examined by electron microscopy revealed particles of a size compatible with filovirions. Indirect fluorescent antibody tests indicated transient seroconversion to Marburg virus. In lymphocyte transformation assays of cells isolated from the patient 11 months after the onset of acute disease, Marburg viral antigen was able to stimulate lymphocyte proliferation 3.9-fold; however, exhaustive attempts to isolate virus from acute phase blood cultured in vitro or in vivo from guinea pigs and monkeys failed. Data suggest that this patient may have been infected with a filovirus. This case demonstrates the difficulties that may occur in laboratory diagnosis of viral haemorrhagic fevers.