The effect of 'low-dose' corticosteroids (9 mg/kg methylprednisolone), given after skeletal trauma, on the incidence of the fat embolism syndrome and isolated arterial hypoxemia was studied in 42 controls and 40 steroid-treated subjects. Fat embolism occurred in ten controls (23.8%) and one steroid-treated subject (2.5%) (p = 0.01). A further 44 subjects developed isolated hypoxemia. This was severe (PaO2 less than 50 mm Hg) in seven of 32 controls (21.9%) and one of 39 steroid-treated subjects (2.6%) (p = 0.01). The overall incidence of hypoxemia was 67.1%, affecting 33 controls (78.6%) and 22 steroid-treated patients (55%) (p less than 0.05). The degree of hypoxemia was severe (PaO2 less than 50 mm Hg) in 12 controls (28.6%) and two (5%) of the steroid-treated subjects (p = 0.005). No control subject died or required mechanical ventilation. One steroid-treated subject without fat embolism died of a fulminant infection. Although methylprednisolone in a relatively low dose provides protection against fat embolism and pulmonary dysfunction after skeletal trauma, the safety of this therapy requires further evaluation.