We report a previously undescribed autosomal reciprocal translocation, t(1;8)(q41;q23.1). It segregates in three families whose common origin lies at least 11 generations back. No examples of unbalanced karyotypes were encountered. Moreover, there was no circumstantial evidence that such live births had occurred during earlier generations. Couples in which one spouse was a translocation carrier were compared to related couples with normal karyotypes. The 15 carrier families had significantly more spontaneous abortions (32%) than the 22 normal couples (10%), irrespective of the sex of the carrier parent. However, the mean number of children was equal in both groups (2.0 and 2.4). Carrier families produced 17 children with a balanced translocation and seven with a normal karyotype. This deviates significantly (P = .04) from the expected 1:1 ratio. We conclude that this malsegregation helps to maintain the translocation in the population. These results show that empirically derived 1:1 segregation ratios previously reported in series that combine many different translocations do not apply to all individual translocations.