Being a putative predictor of ischemic heart disease, the measurement of factor VII (FVII) coagulant activity will be presumably requested to clinical laboratories with increasing frequency. To assess the influence on FVII assays of different thromboplastins and FVII-deficient plasmas we compared performances of all possible combinations of 5 thromboplastins and 6 deficient plasmas. The reproducibility of the clotting times of the dose-response curves for human and rabbit thromboplastins were acceptable (CV lower than 7%), whereas bovine thromboplastin had a higher CV. Reproducibility was very similar for all deficient plasmas when they were used in combination with a given thromboplastin. Responsiveness of the dose-response curve did not depend on the deficient plasma but rather on the thromboplastin: one rabbit thromboplastin was the least responsive, the bovine thromboplastin the most responsive, the human and the remaining two rabbit thromboplastins had intermediate responsiveness. Assay sensitivity to cold-activated FVII varied according to the thromboplastin: the bovine thromboplastin was the most sensitive, the human thromboplastin the least sensitive, of the three rabbit thromboplastins two were relatively sensitive, one was almost insensitive. In conclusion, our results indicate that thromboplastin rather than deficient plasma is the crucial factor in the standardization of FVII assay.