In order to detect a higher recurrence risk for age-dependent chromosome abnormalities, four different types of material are considered. The diíferent types are characterized by their amount of information about the families. For each type of material is given a statistical method to test the hypothesis that the probability for another affected individual is equal to the population incidence against the alternative that the probability is larger in the material. The methods are illustrated by a material on trisomy mongolism for children born by women below 30years and above 30 years. It is detected that there is a recurrence risk higher than the corresponding population incidence for children of women who are below 30 years and who have a trisomy mongol. For other age groups a corresponding increased risk is not detected.