Abstract
For sibships ascertained through a twin-pair where at least one of the twins is affected, an exact conditional test is developed for the hypothesis that the probability for a dizygotic twin to be affected is equal to that of single-born children against the alternative that the probability is higher among twins. Such a situation may occur in disorders where a twin pregnancy may cause sporadic cases of non-genetic origin. The method has been applied to data on febrile convulsions. It is shown that dizygotic twins with a family history of the disorder are more likely to get febrile convulsions than their single-born sibs. I want to thank DR M. Lennox-Buchthal, M.D., Copenhagen, for providing me with data and histories.