International Normalised Ratio (INR) is a standardised method of reporting prothrombin time which compensates for the difference in responsiveness of different thromboplastin reagents used in the prothrombin time assay. The INR is calculated by raising the prothrombin time ratio (prothrombin time of patient over mean prothrombin time of normal subjects) to the power of the International Sensitivity Index (ISI) of the thromboplastin reagent. The ISI is the gradient of a logarithmic regression between the specific thromboplastin reagent in use and an international reference thromboplastin preparation (IRP). Although good inter-laboratory precision of INR values could be achieved in laboratories using manual prothrombin time assays,' 2 several studies have revealed that considerable inaccuracy in the INR values may occur when the prothrombin time assays are performed on automated coagulation instruments.34 In an effort to minimise the variable effects of thromboplastin and instrument interaction caused by the use of automated coagulation analyser systems in clinical laboratories, major suppliers of commercial thromoboplastins in North America now supply user laboratories with an instrument specific ISI (IS-ISI) value for INR calculation. Unlike the "true" ISI value of a thromboplastin, which requires full calibration according to the World Health Organisation (WHO) protocol using the relevant IRP with the manual technique, the IS-ISI value is derived from thromboplastin calibration against an IRP using a prototype instrument, similar in design to the one used by the customer's laboratory.56 However, the potential benefit of this approach is still questionable78 and the degree of INR variability induced by laboratory automation on the clinical care of patients undergoing warfarin treatment remains uncertain.9The purpose of the present study was to examine the magnitude of thromboplastin and coagulometer interactions on the precision of INR values, when the manufacturers' recommended IS-ISI values are adopted for the INR calculation. We must emphasise, however, that because only one instrument of each model of coagulometer was used, our study cannot account for the variable effects of thromboplastin ISIs on individual instruments. In this study five commercial thromboplastins and four models of photo-optical coagulometers frequently used in clinical laboratories in North America were chosen for evaluation.