There are no validated guidelines for administering or monitoring oral anticoagulant therapy in pediatric patients. A pediatric thromboembolism program at the Hospital for Sick Children, Toronto, prospectively monitored consecutive children requiring warfarin over an 18 month period. A uniform protocol was followed and dose adjustments based upon international normalized ratios (INRs). One hundred and fifteen consecutive children; 68 males and 47 females, received warfarin. The age distribution was: <1y (19); 1-5 ys (33); 6-10 ys (20); 11-18 ys (43). Warfarin was used for secondary prevention of venous thromboembolism (n = 56) and primary prevention of thromboembolism (n = 59). Underlying disorders included: congenital heart disease (CHD) without mechanical valves (MV) (49); CHD with MV (18); cancer (8); longterm total parenteral nutrition (7); renal disorders (10); other (23). Treatment length was considered as short term (3-6 mths) n = 37 (32%); longterm (> 6 mths) n = 38 (33%); and life-long n = 40 (35%) of children. While receiving warfarin, 95 children received concurrent longterm treatment with other drugs: 1 drug (28); 2 drugs (27); 3 drugs (21); 4 or more drugs (19). The amounts of warfarin/kg required to achieve INRs of 2 to 3 decreased with increasing age. Children <1 year of age required 0.32 +/- 0.05 mg/kg whereas children 11-18 yrs required 0.09 +/- 0.01 mg/kg; P < 0.001. Monitoring warfarin required an average of 4.0 measurements per month and 1.5 dose changes per month. Changes in warfarin doses were primarily precipitated by drugs, intermittent illness, and changes in diet.(ABSTRACT TRUNCATED AT 250 WORDS)