Context Records collected when sick, injured or dead animals arrive at wildlife care facilities have potential to offer insights into population declines and identify key threatening processes for conservation and management intervention. Aims Records compiled from a centralised Queensland Government database of koala (Phascolarctus cinereus) arrivals to care facilities across South East Queensland were analysed to explore long- and short-term trends in arrivals in terms of seasonality, causes, outcomes and spatial distribution, with a particular focus on insights hospital records could provide into the potential role of disease in koala population declines. Methods Analysis was conducted on over 22 years of records of koalas arriving at care facilities collated by the Queensland Government. We summarised causes of hospital arrivals and outcomes. We utilised time-series methods to explore short-term cyclic dynamics in the data, and spatial tools to document changes in the distribution of koala arrivals. Key results In the long term, koala hospital arrivals increased modestly from 1997 to 2014, before falling into decline by 2018. Long-term changes are dwarfed by short-term fluctuations, including clear annual cyclic dynamics associated with car strike and dog attack, which peak from August to October each year, likely coinciding with the onset of the koala breeding season. Seasonality is also detected in disease-associated arrivals. Known severe declines in wild koala populations in South East Queensland, an area of intensive urbanisation and associated loss of koala habitat, are not reflected in the overall koala hospital arrival numbers. Our analysis suggests that severe local declines in wild koala abundance have been obscured by increases in the catchment areas from which koalas are entering the hospital network. Conclusions Koala hospital records provide an extensive dataset that can be mined for insights into koala population dynamics and threatening processes. However, interpretation of our findings must consider limitations and biases inherent in data collection. Implications Despite acknowledged shortcomings in terms of bias and data quality, retrospective analysis of records from care facilities can provide important insights for guiding conservation efforts. For example, our findings with respect to seasonality in koala hospital arrivals mirror results reported for other locales, suggesting that cyclic dynamics are not a local phenomenon, but occur more broadly across the species range, with implications for seasonal delivery of conservation actions.