Technological innovation has been regarded as the core competence for the economic growth of individual, as well as organization and country. Pharmaceutical innovation, what we call new medicines, influence people`s longevity and productivity by increasing output per hour worked. Therefore, using claims data on virtually all the drugs and diseases of over 550,000 people enrolled in National Health Insurance Program in Korea, we examined the impact of the vintage (original FDA and KFDA approval year) of drugs used to treat a patients from July 1st to December 31st in 2002 on the patient`s mortality at the end of 2004, controlling for demographic characteristics(age and sex), utilization of medical services, and the nature and complexity of illness. We found that people using newer drugs are less likely to die at the end of 2004, conditional on covariates. The estimated mortality rates were declining with respect to drug vintage for 1970s, 1980s and 1990s and highly significant. In addition to estimating the model for the entire sample, we estimated the model separately for several disease categories classified by Korean Classification of Disease. Estimates of three drug vintage variables for subgroups of people with (1)neoplasms, (2)endocrine, nutritional and metabolic diseases, and (3)the diseases of circulatory system displayed similar patterns.