A questionnaire-based, retrospective case-control study of 174 Swedish cats was used to identify possible risk factors for feline Borna disease. The questionnaire consisted of 32 questions on a wide range of subjects, including lifestyle and possible modes of virus transmission. Two control groups were used: a healthy-control and a hospital-control. Variables with significant odds ratios in either the Borna disease:healthy-control or Borna disease:hospital-control comparison were included in multiple logistic regression analyses. Overall, the models suggested that feline Borna disease has a predominantly rural/woodland distribution, that affected cats were more likely to be males than females and intact than neutered, and that they were more likely than not to have hunted mice. The results indicate that, in contrast to other feline viral infections, Borna disease virus is not readily transmitted between cats. The natural reservoir of the virus is unknown. The fact that exposure to mice by hunting was a risk factor for the disease suggests that rodents may be subclinically infected and act as virus carriers.