Background: This study is to evaluate the effect of systemic corticosteroid on the clinical outcomes and the occurrence of complications in mechanical ventilated patients with severe community-acquired pneumonia (CAP). Methods: We retrospectively assessed the clinical outcomes and complications in patients with severe CAP admitted to ICU between March 1, 2003 and July 28, 2009. Outcomes were measured by hospital mortality after ICU admission, duration of mechanical ventilation (MV), ICU, and hospital stay. Complications such as ventilator associated pneumonia (VAP), catheter related-blood stream infection (CR-BSI), and upper gastrointestinal (UGI) bleeding during ICU stay were assessed. Results: Of the 93 patients, 36 patients received corticosteroids over 7 days while 57 patients did not receive corticosteroids. Age, underlying disease, APACHE II, PSI score, and use of vasopressor were not different between two groups. In-hospital mortality was 30.5% in the steroid group and 36.8% in the non-steroid group (p>0.05). The major complications such as VAP, CR-BSI and UGI bleeding was significantly higher in the steroid group than in the non-steroid group (19.4% vs. 7%, p<0.05). The use of steroids and the duration of ICU stay were significantly associated with the development of major complications during ones ICU stay (p<0.05). Conclusion: Systemic corticosteroid in patients with severe CAP requiring mechanical ventilation may have no beneficial effect on clinical outcomes like duration of ICU stay and in-hospital mortality but may contribute to the development of ICU acquired complications.