In order to investigate the possible relationship between physical health and political violence in Northern Ireland a random sample of residents of four electoral areas (two with relatively high violence and two with relatively low violence) was interviewed at home. Each person was asked to rate their health in terms of common physical symptoms, to indicate their use of family doctor and hospital services, and to rate the level of political violence in their neighbourhood. Analysis of covariance (with a measure of psychological well-being, a measure of trait neuroticism plus age and socioeconomic status as covariates) revealed that women reported more physical symptoms than did men, people in the 'high' violence areas reported more symptoms than did those in the 'low' violence areas, while those who rated their own neighbourhood most highly in terms of perceived violence also reported the greatest number of physical symptoms. However, a series of chi 2 tests revealed no association between political violence or perceived political violence and uptake of services.