The influence of three types of farm management on the occurrence of Q fever in cattle was investigated by means of a serological study carried out in Irpinia in southern Italy. Twenty-one herds were permanently housed, 26 were housed in winter and turned out to graze in the spring, and six herds were kept outdoors throughout the year. Blood samples were taken from 1188 cattle and tested by an indirect immunofluorescence assay. The overall seroprevalence of Q fever was 14.4 per cent. The herds which were housed in winter and turned out in the spring had the highest seroprevalence of 19.6 per cent, followed by the permanently housed animals with a seroprevalence of 13.2 per cent and the unhoused cattle had a much lower seroprevalence of 1.9 per cent.