We investigated whether primary (< 24 hours) intramedullary stabilization of femoral shaft fractures in multiple trauma patients with severe thoracic injury might be associated with an increased incidence of adult respiratory distress syndrome (ARDS). A total of 766 patients with multiple trauma admitted to Hannover Medical School between January 1, 1982, and December 31, 1991, were investigated retrospectively. Of these, 106 patients met the inclusion criteria: Injury Severity Score > 18, femoral midshaft fracture treated by intramedullary nailing, primary admission or referral within 8 hours after injury, and no death from head injury or hemorrhagic shock. Two groups were differentiated according to the presence or absence of chest trauma (severe chest trauma = AIS thorax > or =, group T; no severe chest trauma = AIS thorax < 2, group N). Selection of subgroups according to the time of femur stabilization was group I < 24 hours after trauma, group II > 24 hours after trauma. Injury Severity Scores in the four groups were TI: 29.4 (n = 24); TII 31.4 (n = 26); NI 20.1 (n = 33); NII 25.4 (n = 23). In patients without thoracic trauma the ICU time (NI: 7.3 days; NII: 18.0 days) and intubation time (NI: 5.5 days; NII: 11.0 days) were lower in the patients treated primarily (p < 0.05). In patients with severe chest trauma there was a higher incidence of posttraumatic ARDS (33% versus 7.7%) and mortality (21% versus 4%) when early intramedullary femoral nailing was done. In the absence of severe chest trauma primary intramedullary femoral nailing is beneficial.(ABSTRACT TRUNCATED AT 250 WORDS)