Part A. Health Behavior Objectives: The prevention of cognitive impairment, a prodrome of dementia, is crucial due to the increasing life expectancy and growing prevalence of dementia. Among the risk factors of cognitive impairment, health behaviors such as smoking, drinking, sleeping, diet, and physical activity, are potentially modifiable as far as public health intervention is concerned. This study aimed to examine the relation between cognitive function and health behaviors. Methods: Longitudinal data were drawn from “The Survey of Health and Living Status of the Near Elderly and Elderly in Taiwan” in 1993, 1996, 1999, and 2003. Four-year follow-up retrieved from 1999 to 2003, 6-year follow-up 1993-1999, and 10-year follow-up 1993-2003 and the sample size was 1557, 1680, and 1250 respectively. Cognitive function was measured by the Short Portable Mental Status Questionnaire (SPMSQ). SPMSQ contains 9 items and the sum ranges from 0-9 as scores greater than 6 defined as normal. Results: The hazard ratio (HR) of inadequate sleep duration in 6-year follow-up and 10-year follow up were 1.69 and 1.77, respectively. The HR for combination of no alcohol and inadequate sleep duration were 1.61 and 1.80, respectively. The less the physical activity is, the worse the cognitive function is. In addition, the number and the duration of unhealthy behaviors are negatively related to cognitive function. Conclusion: A dose-response relationship between unhealthy behaviors and cognitive impairment was evident. Reducing the number or the duration of unhealthy behaviors might be associated with a reduced risk of cognitive impairment in the elderly. Part B. Leisure activity Objectives: The prevention of cognitive impairment, a prodrome of dementia, is crucial due to the increasing life expectancy and growing prevalence of dementia. Among the risk factors of cognitive impairment, health behaviors such as leisure activities, are potentially modifiable as far as public health intervention is concerned. This study aimed to examine the relation between cognitive function and leisure activities. Methods: Longitudinal data were drawn from “The Survey of Health and Living Status of the Near Elderly and Elderly in Taiwan” in 1993, 1996, 1999, and 2003. Four-year follow-up retrieved from 1999 to 2003, 6-year follow-up 1993-1999, and 10-year follow-up 1993-2003 and the sample size was 1557, 1680, and 1250 respectively. Cognitive function was measured by the Short Portable Mental Status Questionnaire (SPMSQ). SPMSQ contains 9 items and the sum ranges from 0-9 as scores greater than 6 defined as normal. Results: In 6-year and 10-year follow-ups, the hazard ratio (HR) of not playing games/cards was 6.43 and 3.82, and the population-attributable risk percent (PAR%) was 82.59% and 70.89% , respectively. Regardless of length of follow-ups, the less or the shorter duration the leisure activities were, the greater risk the cognitive impairment was. Conclusion: Increasing the number and the duration of leisure activities, especially playing games/cards, was associated with reduced risk of cognitive impairment. Leisure activities might be “cognitive reserve” and further prevent cognitive impairment among the elderly.