Off-label prescribing occurs when a practitioner prescribes a drug for a use, or in a manner, not listed in the 'approved product information' (API) for that drug. The literature suggests that this is a frequent occurrence in many areas of medicine, but data are limited in the hospitalized oncology setting. The aim of this study was to quantify the extent of off-label prescribing in a hospitalized oncology population in Australia. The study was conducted at Peter MacCallum Cancer Centre, Australia. On a single day the medication charts of all hospitalized patients were prospectively reviewed. Drug prescribing was assessed for licensing status by comparison with the API as approved by the Therapeutic Goods Administration of Australia. Prescriptions were classified as licensed, off-label or unlicensed. Medication charts of 130 patients were assessed. There were 1351 prescriptions. In 293 (22%) of the prescriptions the drug was either off-label (242, 18%) or unlicensed (51, 4%). Among the 130 patients, 110 (85%) received at least one drug that was prescribed off-label or that was unlicensed. Off-label dosing was the most frequent reason for a drug being off-label (139, 10% of all prescriptions). Off-label due to use for an unapproved indication was found in 118 prescriptions (9%), and off-label due to an unapproved route of administration was found in 38 prescriptions (3%). Off-label prescribing is widespread in the acute hospitalized oncology population, with approximately 22% of all prescriptions being for off-label or unlicensed medication. Such prescribing affects a significant proportion of patients.