Background: The survival of lung cancer patients in critical care has been improved. Increasing number of lung cancer patients have signed predefined do-not-intubation (DNI) order before admission to ICU. These patients may still be transferred to ICU and even receive non-invasive ventilation (NIV) supports. However, there is still lack of prognostic prediction in this cohort. Whether patients will be benefit of ICU care remain unclear. Methods: We retrospectively collected data of patients with advanced lung cancer, having signed DNI order before ICU admission in a tertiary medical center. The clinical characteristics and survival outcomes were discussed. The enrollment duration is between 2014 to 2016. Results: Total 140 patients were included. All patients were diagnosed with stage III or IV non-small cell lung cancer (AJCC 7^th edition) and signed DNI. The median age was 73 years old and about two-thirds were male. Most patients had received NIV during ICU course. The median APACHE II score was 14 (SE, 0.66) and the mean P/F ratio was 174.2(+/-104) mmHg. The APACHE II score was significantly lower in 28-day survivors (survivor 12 (+/-0.98) vs non-survivor 15 (+/-0.83), p = 0.019). The P/F ratio of the survivors was higher than non-survivors (survivor 209.6+/-111.4 vs non-survivor 157.9+/-96.7, p = 0.006). Patients with P/F ratio ≥150 had better 28-day survival (p = 0.005). By combining P/F ratio≥150 and APACHE II score < 16, those with high P/F ratio and low APACHE II score during ICU admission had a notable 28-day survival compared with the rest (p < 0.001). These prognostic factors could also be applied for 90-day survival (p = 0.003). The prediction model was significant in those with driver mutations in 90-day survival (p = 0.021) but not in those without driver mutations (p = 0.061). Conclusions: The P/F ratio≥150 and APACHE II score < 16 were the prognostic significance of critically ill lung cancer patients with DNI. The prediction could be applied for 90-day survival in patients with driver mutation. The finding was informative for clinical practice and decision-making.