SummaryA study was carried out in order to evaluate the Astrup and Mullertz fibrin plate method for estimating plasminogen activators.Choice of a suitable fibrinogen substrate was found to be the most important factor in setting up a workable assay. Many preparations contained a large proportion of non-clottable protein and plates made from these fibrinogens were usually unreliable. In addition, plasminogen content varied appreciably between preparations so that sensitivity of the method required careful calibration with each new batch of fibrinogen.The effect of additives in the fibrin plate was considered and it was found that calcium chloride alone was sufficient to ensure a stabilised plate which could be stored at 4° C for some time. The addition of tranexamic acid (AMCHA) was found to be a slightly more convenient way of estimating direct proteolytic activity, compared with the traditional heated plate. However neither method distinguished completely between proteolysis and plasminogen activation.In order to improve the precision of the method, the use of an analysis of variance technique has been studied. This technique provides information on the dose-response curves of test and unknown substances, and in addition produces an approximately threefold increase in precision over single plate tests.