目標：探討中老年人其社會參與對存活的影響，並進一步針對喪偶者與有偶者進行比較。方法：資料來源為1999年「台灣地區中老年身心社會生活狀況長期追蹤調查系列研究調查資料庫」，社會參與定義為活動參與及有酬工作，影響存活主要因素為喪偶情形及社會參與，採用Cox迴歸進行分析預測1999至2003年四年之間及1999至2007年八年之間的死亡風險。結果：在四年期間存活分析結果發現持續喪偶者的死亡風險為持續有偶者的1.47倍；65至74歲的老人的死亡風險為53至64歲的2.73倍，75歲以上的老人則5.62倍；認知功能障礙的死亡風險為正常1.34倍；日常生活活動障礙的死亡風險為正常1.76倍；工具性日常生活活動障礙的死亡風險為正常1.99倍。在八年期間存活分析結果發現持續喪偶者的死亡風險為持續有偶者的1.2倍；而男性、年齡較大、非外省籍、罹患三種以上慢性病、有日常活動功能障礙者、自覺健康狀態較差者及有憂鬱傾向者死亡風險較高。結論：社會參與及喪偶對存活有各自的影響作用，喪偶對存活有負向的影響，活動參與對喪偶者及有偶者有正向的作用，有酬工作只對有偶者有正向影響，故社會參與有助於提升中老年人晚年福祉。

Objectives: This longitudinal study examines the effect of social participation on mortality among the elderly in Taiwan. Comparisons were conducted between the widowed and the married survey respondents. Methods: Data were taken from 1999, 2003, and 2007 waves of the survey. Baseline social participation was measured by items asking if the respondents were involved in various social activities and paid jobs. Cox regression models were used to examine if widowhood and social participation were predictive of their mortality over a four-year (1999-2003) and an eight-year (1999-2007) period. Results: In the four-year mortality analysis, it was found that being a widow was 1.47 times more likely to die, comparing to the married. For the mortality analysis between 1999 and 2007, being a widow was 1.2 times more likely to die than the married. Additionally, people who were male, older, and non-provincial, having more than three chronic diseases, with more activities of daily living impairments, with a poorer self-rated health, and feeling depressed had a higher relative hazard of deaths over the study periods. Conclusions: Both social participation and widowhood are significantly associated with mortality. The widowhood had influence on mortality. Improvement in social participation would improve older people's welfare.