SummaryIn contrast to all previous reports of psychoses in genetically identical siblings, identical triplets aged 28 were discordant by major Kraepelinian type of psychosis. Two of them had been diagnosed schizophrenic while the third was diagnosed manic depressive. Re-evaluation of their illnesses using standardized methods and diagnosis by blind raters suggested that the discordance was not due to simple misdiagnosis or different diagnostic bias between psychiatrists. These triplets illustrate some of the shortcomings of a strictly applied Kraepelinian dichotomy and show that a genetic perspective is important in attempting to resolve the classificatory problems of atypical psychoses.