Cryopreservation of sperm cells (cryobank) is a strategy used by some aquaculture and livestock breeding programs to conserve and, if necessary, recover the genetic diversity of a population. The effectiveness of this strategy depends primarily on the methodology of cryopreservation, the viability of mass artificial fertilization with cryopreserved milt/semen, and the selection and culling decisions of the individuals to have their sperm stored, given constraints imposed by storage capacity and related costs. Ideally, the population represented in the cryobank is as genetically diverse as possible and, in some cases, present relatively superior genetic merit. The latter for cryobanks that also have the aim of serving as a backup population of elite broodstock. Different methods to assess and optimize the genetic diversity of a population can be used to assist in determining the storage and management policy of cryobanks. If genotypic information is available, genetic diversity can be optimized at the haplotype level, maximizing the count of unique haplotypes of a given subset of individuals. Modern imputation methods can also have a role in estimating the diversity held within ungenotyped individuals. The objectives of this study were to assess the genetic diversity of the Saltas (Salmon Enterprises of Tasmania Pty Ltd) cryobank and develop a strategy to optimize decision making on which samples to be stored. The Saltas cryobank was primarily developed to preserve the genetic diversity of the Saltas Atlantic salmon selective breeding program (SBP) but subsequently has also been strategically designed to serve as a backup source of nucleus sires. Currently, it stores milt from over 300 sires, from twelve different year classes (2007-2018), in approximately 2,000 Squarepacks® (Cryogenetics Inc). Its storage policy has been based mainly on information from pedigree and estimated breeding values. As 50K SNP genotype information is available for a large proportion of sires in the cryobank and for all broodstock candidates, the genetic diversity in the present study was assessed by computing the number of unique haplotypes from genomic regions (~1,500 windows) comprised by 30 consecutive SNP markers. The results indicated that the Saltas cryobank is currently storing approximately half of the haplotypes segregating in the SBP population. Cryobank samples and broodstock candidates were ranked according to their contribution for the genetic diversity of the SBP population, allowing the development of an optimized approach for future storage policy. The haplotype analyses of available broodstock candidates revealed no loss of genetic diversity in the most recent generations and, consequently, no current need to use cryobank material to recover genetic diversity. This result can primarily be explained by the size of the population (around 200 families produced annually) and by the optimum contribution and mate allocation strategies implemented in the SBP, which maximize genetic gain under constrained levels of coancestry and inbreeding. Opportunities to combine haplotype diversity and optimum contribution theory, for cryobank storage optimization purpose, were envisioned and will be further explored.