Spring radiant frosts occurring when wheat is in reproductive developmental stages can result in catastrophic yield lost for producers. In wheat, heading time is the main determinant to minimize frost risks and to adapt new frost-tolerant cultivars to target population environments. Gene-based phenology models provide robust tools to predict heading times based on alleles of VRN and PPD genes, and have been widely validated across Australian wheatbelt for most of commercial wheat cultivars. A field experiment was conducted at Gatton in 2014 to calibrate the gene-based model for newly released cultivars. The results indicated that one field experiment including extended photoperiod and pre-vernalization treatments can be used to parameterize new cultivars and allow accurate prediction of heading time across all Australian environments using our gene-based model. Across Australia, we found that yield could be improved by up to 20% on average if frost tolerant lines were available. The yield increase resulted from (1) reduced frost damage and (2) the ability to use earlier sowing dates. Simulations suggest that a small reduction in the threshold temperatures, equivalent to frost tolerance of 1°C lower than current cultivars, would have a large effect in the west of Australia. In the east, frost tolerance to lower temperatures (~ −4°C) would be required to maximise the yield advantage.