背景和目的：過去許多文獻均表示自殺率可反應該地區人民的心理健康狀態，失業率可反應該國家的經濟狀況，近年來台灣這兩項時間序列數據的相互影響與衝擊程度為何，是研究者欲討論的重點。 方法：本文擬將變數間之前期不均衡項(period disequilibrium term) 引入向量自我迴歸模型（Vector Autoregressive Model，簡稱VAR）探討國內自殺率與國民失業率之互動與衝擊持續性，並探討其所可能隱含的意義，資料期間為1995年1月至2004年12月的月資料。 結果與結論：本研究實證獲得結果為：該十年間，台灣國民失業率影響國民自殺率的情況在統計上呈顯著關係(Prob=0.0002)，在衝擊反應分析方面，自殺率在面對失業率的變動，呈發散情況，且受衝擊幅度很大，期間很長；而依預測誤差變異數分解的結果來看，台、澎、金、馬地區失業率的變動受自殺率變動的影響隨期數逐漸升高。本研究以時間序列的統計方式呈現出自殺率與失業率間的高互動關係，未來可針對類似的心理衛生議題做更深入的探討，提供包括職能治療師在內的心理健康專業人員做參考。

Background and purpose: Many research suggest that suicide rate could express people's mental health situation, and unemployment rate could express the economic situation of a country. Methods: We introduce the period disequilibrium term amongst variables into Vector Autoregressive model (VAR) to study the suicide rate and unemployment rate co-relationship and the benchmark of the predicable suicide rate. The period of monthly data collected is from January 1995 to December 2004. Results and Conclusion: The empirical results suggest that in the last decade, the unemployment rate influences the suicide rate presents the obvious relation (prob=.0002) and the shock persistence result that is subjected very greatly and the duration very long, also on the predict error covariance decomposition, the fluctuation of the unemployment rate is influenced by the suicide rate fluctuation with expect to go up gradually.