Background. This study evaluates the impact of free annual health examinations on survival of elderly ( :65 years of age) residents in Kaohsiung City, Taiwan. Methods. A stratified random sample scheme was used in each of the 11 districts of Kaohsiung City. A total of 1,193 elderly people were selected and interviewed in 1993; deaths and results of health check-ups were recorded through 1998. Results. While over 50% of the subjects received at least one health examination between 1993 and 1998, only 18% received three or more. Most (60%) subjects who received examinations in a given year also received examinations the subsequent year; most (over 70%) who did not receive examinations in a given year did not receive check-ups the following year. Cox proportional hazards model showed that those who utilized the examination service had better survival probability than those who did not, given the same age, sex, education, marital status, living arrangements, and number of chronic diseases at baseline: The relative risk (RR) of mortality for those who ever utilized the health examination service was 0.50 (P < 0.0001). Conclusions. Elderly subjects who received annual health examinations had lower mortality than those who did not. This finding should be interpreted cautiously, however, as the difference in survival may reflect better general health behaviors among those who participated in the program. (C) 2002 American Health Foundation and Elsevier Science (USA).