Objective: Drugs used in severe status failed in previous treatment are necessary for small target patients even though they are expensive to cover. However, it's difficult to assess whether they should be listed on reimbursement, because most have little evidence including clinical trial and economic evaluation. Therefore we analyzed previous submissions to find out characteristics of those drugs. Methods: We investigated characteristics of 13 submissions of no alternative drugs for reimbursement decision from 2003 to 2007 in Korea and those drugs' reimbursement decisions in other countries. Results: Three drugs (23%) had high level evidence based on SR and RCT, and 11 drugs were recommended in clinical guidelines. Eight economic evaluations (62%) were submitted. Mean additional survival duration was 5.35 month (range 0.8 to 11.1) when life year gained is presented as outcome value (N=7, 54%). Two of them compared with best supportive care or placebo showed ICER was 1.10 GDP/ LYG (0.57 to 1.64) and the others compared with standard therapy showed lower value of ICER (0.86 GDP/LYG, -0.09 to 2.07). Average net amount of budget impact was 9 million USD (431 thousand to 44 million USD). Drugs compared with BSC were almost decided not to be reimbursed except for one drug having longer additional survival, less ICER and lower budget impact. On the contrary, those compared with standard therapy were reimbursed except for one drug having higher ICER(2.07 GDP/LYG) and budget impact. Some drug having low efficiency but low budget impact was listed in Korea. Decisions for reimbursement about those drugs were not consistent among countries (Austrailia, UK, Canada and Korea). Conclusion: Additional life year gained, efficiency and budget impact seemed to affect reimbursement decision in Korea. Further investigation is required for setting of criteria in reimbursement evaluation of no alternative drugs, especially having good cost-effectiveness and high budget impact or the opposite.