Objective To study the mortality and prognostic factors of post-operative acute respiratory failure in cancer patients. Methods There were 1632 postoperative cancer patients from2004 to 2006 in the ICU of Cancer Hospital, in which 447 patients were complicated with acute respiratory failure ( intubation or tracheotomy and mechanical ventilation) . The clinical data was retrospectively analyzed. Stepwise logistic regression analysis was used to identify variables associated with mortality for acute respiratory failure. Results In 447 patients with acute respiratory failure ( male 260, female 187) , 106 cases died with a mortality of 6. 5% . Single factor analysis showed that acute morbodities ( shock, infection, organ failure) , intervention ( continuous renal replacement therapy, vasopressor drugs) , the 28-day ICU free days and APACHE scores ( ≥ 20) had significant differences between the survivor and non-survivor. Multiple logistic regression analysis showed that duration of operation( P = 0. 008, OR 1. 032, 95% CI 1. 008-1. 057) , APACHEⅡ≥20 scores( P =0. 000, OR12. 200, 95% CI 2. 896-51. 406) , organ function failure( P =0. 000, OR 13. 344,95% CI 3. 791-7. 395) were associated with mortality of acute respiratory failure. Conclusion Duration of operation, organ function failure, and APACHE Ⅱ scores were risk prognostic factors for postoperative cancer patients with acute respiratory failure.