41 Background: Currently, palliative chemotherapy (PC) near the end of life is frequently discussed. It is important to know when it is appropriate to end PC to maintain the best quality of life. The criteria for discontinuation of anti-cancer therapy are often “exacerbation of patient’s condition,” “problems with toxicity,” or “patient refusal.” However, there is no clear agreement regarding when to end PC. Ending treatment is one of the most difficult decisions for oncologists. The purpose of this study was to determine clear reasons for the discontinuation of PC and to gain an understanding of the current situation regarding explanation of prognosis in clinical practice. Methods: This retrospective study included 144 patients with incurable cancer who had received PC and died between January 2014 and November 2016. We examined the reasons for discontinuation of PC and whether patients had completed end-of-life discussions, including a discussion of life expectancy, when their PC was terminated. Results: Causes of discontinuation of PC were as follows: 8.3% patient’s desire, progressive disease (PD) in 48.5%, poor performance status (PS) in 20.1%, and toxicity in 17.4%. In patients who chose to end anticancer therapy voluntarily, all patients received a clear explanation of their prognosis, including life expectancy. Among 67 patients with PD, 45 (67.1%) received information regarding their prognosis; 17 of 29 patients (58.6%) with poor PS received similar information. Among patients who ended PC due to toxicity, only 8 of 25 patients (32%) completed end-of-life discussions, including a discussion of life expectancy. In addition, death within 1 month from the discontinuation of PC occurred in 0 cases where treatment was discontinued due to patient’s desire, 10 cases (15%) due to PD, 10 cases (34.4%) due to PS, and 7 cases (28%) due to toxicity. Conclusions: This study shows that patients who discontinued treatment due to poor PS or toxicity had a shorter interval between the final PC and death. Moreover, it indicated that the details of the end-of-life discussion might be influenced by the reason for discontinuation of PC, including or excluding a discussion of life expectancy.