Whether to incorporate animals with unknown ancestries as founders into scientifically managed captive breeding programs, can be a difficult decision. If the animals are offspring of known founders, their inclusion in the breeding program will result in an increased incidence of inbreeding in the captive population. If the animals are additional founders, excluding them from the breeding program will result in the loss of valuable genetic variation. In general, the practice in scientifically managed captive breeding programs is to exclude animals with unknown ancestries to avoid possible inbreeding. A method of estimating the cost of making an incorrect decision on whether to use animals of unknown ancestry as founders both in terms of lost genetic variation and increased inbreeding is presented. It was determined that the loss of genetic variation resulting from excluding founders is always greater than the loss of genetic variation caused by unequal founder line representation resulting from including related animals, as if they were founders. In addition, the increased rate of accumulation of inbreeding resulting from excluding founders will eventually overcome the initial inbreeding resulting from including related animals. However, in some cases, it will take a substantial number of generations for this to occur, and the benefits of possible lowered future expected inbreeding may never be realized. The decision concerning whether to use animals with unknown ancestry should, therefore, be based on the estimated relative costs of making an error, in terms of both lost genetic variation and expected future inbreeding, rather than on avoiding the immediate possibility of increased inbreeding alone. Two examples using studbook data are given to show how this method can be practically applied to the management of captive populations. © 1993 Wiley-Liss, Inc.