AIM: To assess the effect of a lyophilised calibrant plasma procedure on the international normalised ratio (INR) and its interlaboratory variation. METHODS: INR equivalent values were assigned to five lyophilised plasmas (one from normal donors and four from coumarin treated patients) by a reference laboratory using three calibrated thromboplastin reagents. The calibrant plasmas and five artificial control blood specimens were mailed to 44 Dutch laboratories for prothrombin time (PT) determination. The assigned INR values were used to calculate calibration lines for each participant laboratory. The calibration lines were then used to translate the PT of the control specimens to INR. RESULTS: For all lyophilised plasmas and control blood samples, there were significant differences between INR values determined with the three thromboplastin reagents. These differences could not be explained by inaccuracy of the international sensitivity index or mean normal PT of the reagents and must, therefore, have been induced by the preparation procedures for the lyophilised plasmas and control blood samples. The interlaboratory variation of the INR obtained with the calibrant plasma procedure had a coefficient of variation (CV) ranging between 2.1% and 7.3% and tended to be lower than the interlaboratory variation found with the usual methods (3.0-12.2% CV). There was a good agreement between the mean INRs obtained with the calibrant procedure and those obtained using the normal methods. CONCLUSIONS: The present study highlights the limitations of some lyophilised plasmas and control blood samples. It is not possible to assign a single INR value to each of these lyophilised plasmas and control specimens that is valid for all thromboplastin reagents. Nevertheless, by using reagent specific INR equivalent values for the calibrant plasma procedure, the interlaboratory variation could be reduced.