A population in which parentage information is not recorded is considered a pedigree “black hole” in terms of genetic management. Integration of animals from such an unpedigreed population into a genetically managed population is often accomplished by assuming a worst-case scenario and recording all animals as highly interrelated (e.g., full siblings). The assumption that a worst-case scenario is appropriate under a wide variety of circumstances is evaluated. It was determined that the genetic costs of overestimating relationships among animals from an unpedigreed population are always greater than the genetic costs of underestimating relationships by the same amount. This demonstrates that worst-case scenarios, which are by definition intentional overestimates, are not generally appropriate. A method of obtaining an estimate that will yield the lowest overall genetic cost using the possible maximum and minimum levels of relatedness among animals from the unpedigreed population is presented along with two methods for implementation. Finally, for cases in which the estimated average relatedness among animals from an unpedigreed population is low, it is recommended that different analytical studbook datasets be used for calculation of inbreeding coefficients and mean kinship coefficients. This strategy would allow close estimation of average relatedness without risking inadvertent pairing of highly related animals from the unpedigreed population. Zoo Biol 20:305–314, 2001. © 2001 Wiley-Liss, Inc.