The agarose gel method for the detection and quantitation of Factor VIII inhibitors was investigated and compared with the Bethesda method. The agarose gel method was standardized for this study by modifications of the original method of Bird. (Bird P: Coagulation in an agarose gel and its application to the detection and measurement of factor VIII antibodies. Br J Haematol 1975; 29:329-340) Precision studies on values obtained by the agarose gel method indicated it is a reproducible assay. Random error was evident in both methods, but variance appeared to be greater in the Bethesda method for most of the samples evaluated. Comparison of the two methods with split patient samples indicated that proportional error is present in the agarose gel method, and that standardization of the method requires additional study. The agarose gel assay detected low levels of inhibitor (down to 3.4 Bethesda units) and all positive samples were identified. A screening procedure modification of the agarose gel method provided the sensitivity to detect Factor VIII inhibitor levels of less than 1.0 Bethesda units. The results indicated that the agarose gel method is a feasible alternative to the Bethesda method for both quantitative and qualitative determination of Factor VIII inhibitors. Furthermore, since reproducibility is better, it may be useful in detecting the "true" fluctuations of inhibitor titers in a given patient.