Of the wildlife casualties admitted to rehabilitation centres in England, less than half are subsequently released back into the wild. If the factors associated with survival within rehabilitation centres can be determined, they may be used to focus efforts on individuals with high chances of successful recovery, and thus improve welfare by devoting resources to those animals that are more likely to benefit. We analysed the medical record cards of eight species admitted to four wildlife rehabilitation centres run by the Royal Society for the Prevention of Cruelty to Animals between 2000-2004 to determine those factors that affected the chance of survival in care until release, and whether trends in predictive factors occurred across taxonomic groups. We found that the most important predictive factor, across all species, was the severity of the symptoms of injury or illness. Factors commonly used as important indicators of rehabilitation success in published practice guidelines, such as mass and age, were not found to affect survival significantly. Our results highlight the importance of triage based on clinical diagnosis as soon as a wildlife casualty is admitted, and indicate that although the ethos of many rehabilitation centres is to attempt the treatment of all wildlife casualties, the attempted treatment of those with severe injuries may be adversely affecting welfare by prolonging suffering.