A simulation study was used to examine the potential use of DNA fingerprinting (DNA tagging) as a tool to avoid excessive inbreeding by identifying suitable candidate breeders in genetic selection programs. ‘Broodstock fitness’ (the ability of broodstock to survive from harvest and reproduce) needs to be considered in designing breeding programs using DNA tagging. In this study, reduced broodstock fitness increased inbreeding exponentially. The level of inbreeding was also dependent on the intraclass correlation (t), selection intensity, number of individuals DNA tagged (N_DNA), number of families maintained (N_f) and the number of candidate breeders retained per sex/family at harvest (C). With a broodstock fitness of 0.90, DNA tagging could theoretically achieve a selection intensity, in terms of the total phenotypic variance, of 2.90 standard deviations with 800 000 graded at harvest, while maintaining an inbreeding rate of 1.0% per generation (N_DNA = 800, N_f = 30, C = 4, t = 0.3). In practice, the numbers required could be achieved by growing families in individual facilities (e.g. sea cages for barramundi or ponds for prawns). When mechanical grading is not possible, the selection pool may be limited to a level where physical tagging is feasible. In this case, there was no advantage in selection response using DNA tagging compared with physical tags. DNA tagging as a selection tool may be more feasible when broodstock fitness is above 0.6, and may fill a niche where industry infrastructure is not large enough to support separate rearing of families or where physical tagging is not economically viable or suitable. DNA tagging may also be useful as a means of recovering families in backup facilities where families have been pooled to reduce infrastructure costs. Due to the random nature of DNA sampling, not all families may be recovered and a reduction in selection pressure may facilitate family recovery.