本研究目的為瞭解大學生自我分化、壓力感受與身心健康之間的關係。主要想探討一、大學生自我分化、壓力感受與身心健康之狀態；二、大學生自我分化與身心健康的關係；三、壓力感受在自我分化與身心健康關係中的角色。 本研究為問卷調查法，採用「大學生自我分化量表」、「日常瑣事量表」與「一般健康量表」為研究工具，以立意抽樣的方式選取北部地區大學生為研究對象，有效樣本為584名。調查所得資料以描述統計、積差相關與階層迴歸進行統計分析。 本研究主要發現如下： 一、大學生自我分化量表是一具有信效度，可測量個人觀自我分化程度之工具 本研究使用的大學生自我分化量表，是參考Chabot情緒分化量表與自我分化量表修訂版中有關個人觀點自我分化的題目，運用極端組檢定、相關分析法、α係數與因素分析，進行選題與信效度考驗。本量表共有四個因素：「我位置」、「情緒為主」「受他人影響」、「理智為主」，共計27題，具有良好的信、效度。 二、大學生自我分化、壓力感受與身心健康之現況 整體而言，大學生具有中等程度之自我分化、較低之壓力感受與較佳之身心健康。 三、大學生自我分化與身心健康具有顯著負相關 大學生自我分化程度越高，擁有較佳的身心健康。此外，「情緒為主」、「受他人影響」較「我位置」、「理智為主」與身心健康有較高的相關性。 四、大學生壓力感受在自我分化與身心健康的關係上具有有條件的調節作用 本研究以全體樣本探討壓力感受在自我分化與身心健康的關係中的調節效果時發現，壓力感受不具有調節效果。在刪除較低壓力感受50％之樣本再次進行分析，發現壓力感受之調節效果達顯著，而且在不同壓力組中，自我分化對身心健康的解釋力有所不同。 最後，研究者依據研究結果加以討論，並對未來研究及實務工作提出具體建議。 / The purpose of this study was to explore the relationships among self-differentiation, stress, and health of college students in Taiwan. The questionnaire survey was adopted in this study. The research instruments included:“Differentiation of Self Inventory”, “Hassles Scale”, and “General Health scale”. There were 584 college students participants from north district in Taiwan. Data obtained in this study were analyzed by descriptive statistics, Pearson Correlation, and Hierarchical regression analysis. The main results were as follows: 1.The Differentiation of Self Inventory is designed to measure Bowen's intrapsychic aspect of differentiation consisting of “I-Position”, “emotion-based”, “affected by others”, and “intellectuality-based”were of high reliability and validity.. 2.In the average, the college students in Taiwan have medium self-differentiation, lower stress, and better health. 3.In support of Bowen family systems theory greater differentiation of self among college students have better psychological and physiological health. 4.After removing the participants with scores of the Hassles Scales less than 50%,the daily hassles stress was the moderator between self-differentiation and health. In addition, the R square of self-differentiation predicting health was higher in high-stress group than in low-stress group. According to above findings, implications for guidance and parenting were discussed, and suggestions for further research were proposed.