Background: Studies have demonstrated that the risk of warfarin-related complications is highest in the first 90 days of treatment, while quality audits suggest that warfarin initiation protocols are not always adhered to. Aim: To improve the quality of anticoagulation for hospital patients initiated on warfarin. Method: A warfarin drug chart, incorporating the hospital's warfarin initiation protocol, was implemented on four target medical and cardiothoracic wards. The chart was used to record international normalised ratios (INRs) and prescribe warfarin, and was faxed to the patient's GP on discharge. Warfarin initiations on the target wards immediately postintroduction of the warfarin drug chart were reviewed and compared to the pre-intervention data collected over a 17-month period. Results: The pre-intervention and post-intervention groups included 271 and 183 patients initiated on warfarin, respectively. The intervention was associated with fewer warfarin-related complications (thromboembolism, major bleeding) occurring within a 90-day follow-up period (2.1% vs 11%; adjusted OR 0.24; 95%CI 0.07-0.85; p = 0.03), largely driven by a reduction in the incidence of major bleeding (0.7% vs 7.1%; adjusted OR 0.17; 95%CI 0.02-1.45; p = 0.05). Adherence to the warfarin initiation protocol improved significantly from 36% to 71% (p < 0.01) in the pre-and post-intervention data collection periods, respectively. Post-intervention, the proportion of patients with an INR > 4 and the proportion of patients who required a warfarin dose to be withheld was reduced from 8.5% and 13%, to 3.8% (p < 0.05) and 6.6% (p = 0.02), respectively. Conclusion: Adherence to an age and disease based warfarin initiation protocol, and improving the quality of discharge information for patients initiated on warfarin, significantly improved patient outcomes. J Pharm Pract Res 2007; 37: 295-7, 300-2.