Winter birth, urban birth and/or childhood residence, and perinatal complications have each been identified as environmental risk factors for the later development of schizophrenia, schizoaffective disorder, and bipolar disorder. A preliminary case-control study also identified cat exposure in childhood as a possible risk factor. To assess selected environmental events, including childhood exposure to pets, as possible risk factors for these diseases, a case-control telephone survey was carried out by the University of Maryland Survey Research Center for 264 mothers of cases and 528 mothers of matched controls. The cases were randomly selected mothers who were members of the National Alliance for the Mentally Ill, and whose children had been diagnosed with schizophrenia, schizoaffective disorder, or bipolar disorder. The controls were mothers randomly selected from the same telephone exchanges. For five of the 19 major variables, there were statistically significant differences between cases and controls: fever during pregnancy, complications during delivery, city or suburban residence at birth, cat ownership between birth and age 13, and breast-feeding. In a multivariate logistic regression including these five variables, each variable made a significant contribution. The finding of perinatal complications, urban/suburban residence at birth, and cat ownership in childhood as risk factors for the later development of psychoses confirmed previous studies. Previous research on breast-feeding as a risk factor has yielded contradictory results. Additional research is needed to ascertain how such environmental risk factors interact with genetic risk factors. Understanding these could lead to better treatments and possible prevention strategies.