This study aimed to investigate the accuracy of home International Normalized Ratio (INR) self-monitoring in pediatric patients on long-term oral anticoagulation therapy. Statistical and clinical agreement of INR values from capillary whole blood samples measured by 2 different portable prothrombin time monitors (CoaguChek S and XS) and venous blood samples measured by a laboratory coagulation analyzer were evaluated using the Bland-Altman analysis. Eighty-three INR comparisons (56 using the CoaguChek S and 27 using the CoaguChek XS) were obtained from 35 children aged 4 months to 18 years. Mean differences between venous and capillary INR values and their limits of agreement were -0.04 (-0.63 to 0.55) overall, 0.006 (-0.63 to 0.65) for the CoaguChek S and -0.13 (-0.57 to 0.31) for the CoaguChek XS. The Pearson correlation coefficients were 0.88 overall, 0.84 for the CoaguChek S and 0.95 for the CoaguChek XS. Expanded and narrow agreements for all patients were 97.6 and 94%, respectively. In conclusion, home INR self-monitoring is accurate for children requiring long-term oral anticoagulation therapy. Our data suggest that INR self-monitoring with the newer CoaguChek XS is more accurate than with the older CoaguChek S monitor.