The interlaboratory variation of the International Normalized Ratio (INR) in various external quality assessment schemes is still relatively high. This is partly caused by inaccuracy of manufacturers' stated International Sensitivity Index (ISI) and/or local instrumentation effects. The interlaboratory variation and accuracy of INR determinations may be improved by a local calibration procedure based on lyophilized plasmas with assigned INRs. The purpose of the present study was to determine INR values for different types of lyophilized plasmas to be used for local calibration. A total of 13 lyophilized plasmas (one normal, six from coumarin-treated patients, six artificially depleted) were analyzed by 10 laboratories, each using five calibrated prothrombin time (PT) systems. INRs were calculated for each plasma using each laboratory's specific ISI and mean normal prothrombin time values. In the same way, five deep-frozen pooled plasmas from coumarin-treated patients were analyzed. There were significant INR differences for the lyophilized plasmas between the prothrombin time systems. The differences were relatively small for the deep-frozen coumarin plasmas (CV 2.6-3.3%) and three lyophilized coumarin plasmas from one manufacturer (CV 3.7-4.8%). Important INR differences were observed for three lyophilized coumarin plasmas from another manufacturer (CV 9.5-14.1%) and several artificially depleted plasmas (CV 5.3-12.8%). The citrate concentrations in the artificially depleted plasmas were lower than those in the normal and coumarin plasmas. These differences should be considered in the selection and certification of plasmas as calibrants for local calibration of PT systems. The lyophilized plasmas' INR values obtained in the present study will be used for a field study of local PT calibration to assess their efficacy.