Warfarin is frequently administered to hospital patients. The prescription and administration of this medication are particularly susceptible to error. Factors contributing to this include the narrow therapeutic index, patient-specific target range, and the need for regular INR monitoring. NICE guidelines state that warfarin should be given at the same time every day and the Bristol Royal Infirmary guidelines are warfarin to be given at 14:00. The 14:00 dosing ensures standardisation of administration; poor adherence to this recommendation may cause patient harm. We noticed that many warfarin doses were often given outside of maximal staffing hours and it was often left to the on call doctor to prescribe warfarin at erratic and inconsistent times. Our primary aim was to reduce the number of adverse outcomes associated with warfarin prescription and administration. We targeted two system measures: the proportion of warfarin administrations occurring within an hour of the 14:00 prescription and the proportion of INR results outside target range. We employed the model for improvement and carried out our project across seven acute medical wards. Baseline data showed that only 24% of doses were being given within an hour of the recommended time and 64% of doses were being given after 17:00 during minimal staffing hours. We successfully introduced a warfarin box within our trust which demonstrated an improvement in warfarin administration from 24% of patients receiving their warfarin within an hour of 14:00 to 49% and this was subsequently associated with a reduction in INRs above target range (23% to 9%).