The frequency of and reasons for treatment modifications related to prescriptions of antineoplastic drugs and to what extent the modifications are performed in accordance with the local treatment protocol were studied at an oncology day-care unit. Ninety-three patients were treated with antineoplastic drugs at the unit during the study period. Their diagnosis included mainly breast- and gastrointestinal tumours. Thirty-eight treatment modifications in relation to the treatment protocol were observed in 31 of 93 patients (33%). Twenty-five of 31 patients were treated with palliative intention (81%). Two treatment modifications of 38 (5%) were in accordance and 21 modifications (55%) were not in accordance with the local treatment protocol. It was not possible to verify whether the remaining 15 modifications (39%) were according to the protocol. Adverse effects were the most common reason specified in the medical file for treatment modification (8 patients; 26%). The reasons for treatment modification were only documented in the medical file for 11 of 31 patients (35%) and only present on the prescription card delivered to the local pharmacy for one of 31 patients (3%). Drug interactions were not considered according to the medical files for any of the 93 patients who were treated at the unit during the study days, and accordingly, no treatment modifications had been performed due to drug interactions. Liver and/or renal function tests were missing in the medical file for four patients treated with drugs for which these tests are crucial. More emphasis should be put on identifying clinically relevant drug interactions between antineoplastic drugs and the patient's regular drugs and also on specifying the reason for modifications in the medical file and on the prescription cards delivered to the local pharmacy. Increased quality assurance of the local treatment protocols is warranted.