本研究利用自覺團體、個案工作、深度訪談、問卷、以及參與觀察等多種資料蒐集方法，探討臺北市某社區原住民的婚姻暴力狀況，以及受暴婦女的因應措施 , 並評估在社區進行以女性主義充權為方針的個案工作與自覺團體之可行性，最後提出我們對原住民婚暴社區服務模式的建議。主要發現是該社區原住民的婚暴狀況(發生率與暴力型態)或社區居民對婚暴問題的某些迷思，都跟漢人社會相當類似。但社區的人際互動特性以及作為少數族群被主流社會歧視的經驗，卻深刻影響了受暴婦女的因應模式、求助歷程、及協助需求的特性。在社區緊密的人際關係籠罩下，婚暴的傷痛成為受暴婦女不願說、不可說、要自己忍隱的問題。根深的族群歧視則造成社區居民擔心婚暴會繼酗酒、雛妓、失業等問題，成為主流社會給原住民貼上的負面標籤。受暴婦女承受了這種壓力，擔心向正式協助系統求援是丟原住民的臉。她們不願意向警政與社福等機構求助，一方面也是不適應機構式的協助方式，另一方面是因為某些機構專業人員用一套迷思來解釋原住民的婚姻暴力問題，不願意積極加以協助，進一步阻礙了她們的求助意願。最後，本文建議都市或部落原住民社區的婚暴社工處遇模式，宜朝向社區化工作模式，首要關鍵必須先深入瞭解原住民的社會文化脈絡及其社區特有的文化屬性之後，再策劃能夠吻合其文化特性的介入模式。以充權為基礎的自覺團體與諮商輔導原則上適用於原住民婦女，但在操作上必須作些修正以切合個別原住民社區的特有屬性。

　　　　　This study explores the problems of marital violence in urban aboriginal communities relying on data collected from in-depth interviews, questionnaires, and personal observations. Female victims' reactions to violence and evaluation of the viability of casework and of the effectiveness of a consciousness-raising group based on feminist self-empowerment guidelines are also provided, and a community service model regarding aboriginal marital violence problems is proposed. Aboriginal marital violence (frequency of occurrence and type of violence) and its surrounding myths in the subject community were found to be much the same as in the Han society. However, the community's specific social interactions and its encounter with prejudice as a minority in the mainstream society profoundly shape the nature of a victim's reaction to violence and the help-seeking process and needs. In a tightly-knit community, marital violence is a no-tell, can't-tell, keep-to-oneself problem. Because of experiences with deep-rooted racial discrimination, aboriginal communities are concerned that marital violence could become another negative label—like alcoholism, child prostitution, and unemployment—for the mainstream society to attach to aborigines. Under such pressure, victims of violence fear that to seek help from institutions would bring shame to the boriginal community. Moreover, they are unwilling to ask for assistance from police or social welfare institutions both because they are not accustomed to the formalities of institutional process, and because they are discouraged by the indifference of police and social workers who use their preconceptions to explain away aboriginal marital violence problems. We suggest that when handling marital violence problems in urban communities or tribal settlements, social workers should follow a communal model, the key to which is first to understand thoroughly the aboriginal socio-cultural system and the community-pecific cultural features, and then to create a fitting model for intervention. Feminist empowerment consultations and consciousness-raising groups are in principle suitable for aboriginal women, but in practice must be modified to relate to the character of the particular aboriginal community.