OBJECTIVES: The study reported smoking rates in different age groups of the elderly in a Chinese Longitudinal Survey on Healthy Longevity (CLSHL), and illustrated the effects of smoking for the elderly on health status and medical expenditure. METHODS: The CLSHL data was analyzed in this study. CLSHL were cross sectional surveys conducted in 1998, 2000, 2002 and 2005, and participants were restricted to the population aged 65 and older. The range of sample size was from 8374 to 16064 old people from 22 provinces in China. The linear regression model was applied for the analysis of the effects of smoking on health status (self-rated health status and Activity of Daily Living Scale, ADL) and medical expenditure. RESULTS: As age increased, both the current smoking rate and former smoking rate was decreased. For the age groups of 65-79, 80-89, 90-99, 100+ years, there were 26.73%, 8.79%, 7.5%, 6.33%, and 28.30%, 20.02%, 14.46%, 8.51% of smokers in 2002, 2005 respectively.  The former smokers were accounted of 39.21%; 25.26%, 23.07%, 17.99% in 2002, and 40.66%, 36.97%, 30.27%, 20.84% in 2005 respectively. The regression coefficient of self-rated health status in smokers was 0.18, and in former smokers was -0.14. The regression coefficient of ADL in smokers was 0.20, and in former smokers was -0.07.  The regression coefficient of medical expenditure in smokers and former smokers were 0.29 and 0.11 respectively. CONCLUSIONS: Former smoker rates were significantly higher than current smoker rates, which showed that smokers quitted smoking at old age.  Smoking had negative effects on their self-rated health status as well as their ADL. Smoking was significantly related to the medical expenditure of the elderly.