ABSTRACT: BACKGROUND: The provision of surgical assistance in conflict is often associated with care for victims of violence. However, there is an increasing appreciation that surgical care is needed for non-traumatic morbidities. In this paper we report on surgical interventions carried out by Medecins sans Frontieres in Masisi, North Kivu, Democratic Republic of Congo to contribute to the scarce evidence base on surgical needs in conflict. METHODS: We analysed data on all surgical interventions done at Masisi district hospital between September 2007 to December 2009. Types of interventions are described, and logistic regression used to model associations with violence-related injury. RESULTS: 2869 operations were performed on 2441 patients. Obstetric emergencies accounted for over half (675, 57%) of all surgical pathology and infections for another quarter (160, 14%). Trauma-related injuries accounted for only one quarter (681, 24%) of all interventions; among these, 363 (13%) were violence-related. Male gender (adjusted odds ratio (AOR)=20.0, p<0.001), military status (AOR=4.1, p<0.001), and age less than 20 years (AOR=2.1, p<0.001) were associated with violence-related injury. Immediate peri-operative mortality was 0.2%. CONCLUSIONS: In this study, most surgical interventions were unrelated to violent trauma and rather reflected the general surgical needs of a low-income tropical country. Programs in conflict zones in low-income countries need to be prepared to treat both the war-wounded and non-trauma related life-threatening surgical needs of the general population. Given the limited surgical workforce in these areas, training of local staff and task shifting is recommended to support broad availability of essential surgical care. Further studies into the surgical needs of the population are warranted, including population-based surveys to improve program planning and resource allocation and the effectiveness of the humanitarian response.