Heat stress is an issue of growing concern for many livestock production systems worldwide affecting not only animal welfare, but also farm profitability. In December 2017, genomic estimated breeding values for heat tolerance in dairy cattle were released for the first time in Australia. The data set was constructed by merging herd-test production records with weather station data. Heat tolerance phenotypes were defined as the rates of decline in milk, fat and protein yield after a heat stress event (i.e., temperature-humidity index exceeds 60), and were estimated using a reaction norm model. The genomic prediction equation was developed from a reference population of 2,236 sires (with heat tolerance phenotypes on daughters) + 11,853 cows for Holsteins and 506 sires + 4,268 cows for Jerseys. These sires and cows were genotyped with 46,276 SNP. Each component of heat tolerance (genomically predicted decline in fat, protein, and milk) is weighted by its economic value, which is assumed to be the same as their weights in the Australian selection indices. The genomic breeding values are then standardised within breed to have a mean of 100 and standard deviation of 5. Although the reliability of this new trait is moderate (on average around 38%), it is expected that this will improve as the reference populations are increased. The genetic trend for heat tolerance has worsened, which is consistent with the correlation with the Australian national selection index (Balanced Performance Index; BPI) which is −0.20. Heat tolerance is currently not part of the BPI, however, its inclusion will be considered as part of the next review of the index.