Not all dog adoptions are successful. This two-year retrospective study used survival (i.e., time-to-event) analyses to investigate readmissions for dogs adopted from RSPCA Queensland shelters between 1 January 2019 and 31 December 2020. A better understanding of temporal patterns and risk factors associated with readmission may help RSPCA Queensland shelters better target and tailor resources to improve retention by adopters. The failure function (the cumulative percentage of adoptions that were readmitted by day of the adoption period) increased rapidly during the first 14 days of the adoption period. Approximately two-thirds of all returns occurred in this period. This readmission rate may have been influenced by the RSPCA Queensland adoption-fee refund policy. The cumulative percentage of adoptions that were readmitted plateaued at just under 15%. Dog size, age, coat colour, breed, and spending time in foster before adoption were factors associated with the risk of readmission. Failure functions for a low and a high-risk adoption example demonstrated the large degree of difference in hazard of readmission between covariate patterns, with estimated percentages of adoptions being returned by 90 days for those examples being 2% and 17%, respectively. Spending time in foster care before adoption appears to be protective against readmission, presumably because it supports a successful transition to the new home environment. Behaviour support and training provided for dogs during foster care may contribute to improve their outcomes. These findings highlight the profile of the higher-risk dogs potentially providing shelters with an opportunity to examine where and how resources could be allocated to maximize outcomes for the overall cohort. Population attributable 90-day failure estimates were calculated for each of bodyweight and age at adoption, coat colour, spending time in foster care before adoption, and breed. This calculation shows the expected reduction in the cumulative percentage of dogs readmitted by day 90 if the hazards of readmission for higher risk categories were reduced to those of a lower risk category. Expected reductions for individual factors ranged from 1.8% to 3.6% with one additional estimate of 6.8%. Risk of readmission could be reduced through increased development of foster capacity and capability, targeted interventions, improved adopter-dog matching processes, and more effective targeting of support for higher risk dogs, such as older or larger dogs. Population impact analyses provide a macro view that could assist shelters in strategically assessing the return on investment for various strategies aiming to improve adoption outcomes and potentially reduce readmissions.