In comparison with British Comparative Thromboplastin, we examined 11 different reagents which are presently available in Switzerland for surveillance of oral anticoagulation. Frozen plasma samples obtained from 50 patients under oral anticoagulation for at least 6 weeks were used for comparison of the various reagents. From our results the following conclusions emerge: Our present results largely concur with those of a similar study which compared the therapeutic range of thromboplastin reagents available in Switzerland 6 years earlier; therefore, we assume that the reagents which were sold at that time and are now available in Switzerland have remained fairly constant. In contrast to this apparent agreement, the therapeutic ranges indicated by the producers of reagents are, at least in part, comparable neither with the therapeutic range established for the British Comparative Reagent nor among themselves. Thus, the intensity of the induced coagulation defect is obviously variable, depending on a given test system. The calculated "therapeutic range" for a given test system traditionally corresponds to an average risk of thrombosis or bleeding; this range is not of necessity identical with a target zone that has to be established either individually or for certain diseases. Once established, a target value defining the intensity of oral anticoagulation can be more conveniently reproduced by use of suitable control plasmas rather than with reference thromboplastins, at any rate by practitioners or non specialised hospital laboratories.(ABSTRACT TRUNCATED AT 250 WORDS)