Background It has not been investigated how much the use of clinical factors in a dosing algorithm improves the percentage of time in therapeutic range (TTR). The present study aimed to compare the effect of dosing algorithms for acenocoumarol and phenprocoumon including clinical patient characteristics with standard care in the Netherlands. Setting: The pre-EU-PACT study, an observational study in the Netherlands, was used to obtain standard care data. Data from the Dutch patients in the EU-PACT trial (comparing the use of a clinical algorithm with and without genetic information) was used for the clinical dosing algorithm. Methods For both acenocoumarol and phenprocoumon, the percentage of time in, below and above therapeutic International Normalized Ratio (INR) range during 12 weeks after treatment initiation were assessed in both studies. Results During the weeks 2-12, the clinical dosing algorithm of acenocoumarol (80 patients) led to a higher TTR (74.3% versus 68.0% in range 2.0-3.5, 95% Confidence interval [CI] difference: 0.5% to 11.8%), and a reduced percentage of time below INR 2 and above INR 3.5, compared with standard care (272 patients). For phenprocoumon, compared with standard care (484 patients), 80 patients treated by the dosing algorithm did not obtained a significantly higher TTR in range 2.0-3.5 or a lower percentage of time above 3.5, however, they spent more time with INR below 2. Conclusion The use of a clinical dosing algorithm for acenocoumarol seemed to improve the quality of anticoagulation therapy during the treatment of initial 2-12 weeks. For phenprocoumon, there was no statistically difference in anticoagulation control.