Cape Town, South Africa experienced an upsurge in the level of political violence from May to July of 1986. To determine the impact of the political violence on health and health services, selected routinely available information was analyzed, a community survey was conducted of 1,540 randomly selected households in high, medium, and low impact areas (defined using police and community reports), and a survey of 162 nurses (75 per cent response rate) working in clinic and maternity services in Cape Town's townships was undertaken. Methodological problems were encountered in relation to sampling, interviewer allocation to areas, and access to routinely available information. Nevertheless, a consistent picture emerged from the studies that: demonstrated the impact of political violence on attendance at routine health service facilities (for hypertension, tuberculosis, immunizations, antenatal and postnatal services); highlighted the disruptions caused to basic services in high impact areas (water, street lighting, sanitation and transport); documented the problems experienced by nurses in performing their usual services and by patients obtaining access to their services; showed that high impact areas had three times higher rates of gunshot wounds than low impact areas during the period.