Influenza B virus (IBV) strains are one of the components of seasonal influenza vaccines in both trivalent and quadrivalent formulations. The vast majority of these vaccines are produced in embryonated chickens´ eggs. While optimized backbones for vaccine production in eggs exist and are in use for influenza A viruses, no such backbones exist for IBVs, resulting in unpredictable production yields. To generate an optimal vaccine seed virus backbone, we have compiled a panel of 71 IBV strains from 1940 to present day, representing the known temporal and genetic variability of IBV circulating in humans. This panel contains strains from the B/Victoria/2/87-like lineage, B/Yamagata/16/88-like lineage and the ancestral lineage that preceded their split to provide a diverse set that would help to identify a suitable backbone which can be used in combination with hemagglutinin (HA) and neuraminidase (NA) glycoproteins from any IBV strain to be incorporated into the seasonal vaccine. We have characterized and ranked the growth profiles of the 71 IBV strains and the best performing strains were used for co-infection of eggs, followed by serial passaging to select for high-growth reassortant viruses. After serial passaging, we selected 10 clonal isolates based on their growth profiles assessed by hemagglutination and plaque-forming units. We then generated reverse genetics systems for the three clones that performed best in growth curves. The selected backbones were then used to generate different reassortant viruses with HA/NA combinations from high and low titer yielding wild type IBV. When the growth profiles of the recombinant reassortant viruses were tested, the low titer yielding HA/NA viruses with the selected backbones yielded higher titers similar to those from high titer yielding HA/NA combinations. The use of these IBV backbones with improved replication in eggs might increase yields for the influenza B virus components of seasonal influenza virus vaccines.