Summary. Seven 'field' collaborative studies on factor (F)VIII concentrate potency measurements were carried out, using local routine methodology, standards and calculation of results. Data from five of the 12 different concentrates studied are described in detail. These studies revealed that, for the intermediate-purity and the recombinant FVIII concentrates, one-stage potencies were significantly lower than chromogenic potencies, whilst for the two high-purity FVIII concentrates one-stage potencies were significantly greater than chromogenic potencies. On comparing predilution methods for the intermediate-purity concentrate, equivalent potencies were obtained using either buffer or FVIII-deficient plasma as prediluent. For the two highpurity and the recombinant concentrates, potencies obtained using buffer as prediluent were significantly greater and lower, respectively, than potencies obtained using FVIII-deficient plasma as prediluent. Interlaboratory variabilities were compared over all 12 concentrates studied and coefficients of variation (CVs) for one-stage assays were found to be much greater than for chromogenic assays. This was true for all concentrates except for the intermediate-purity concentrate and samples A and B from the first study, where the reverse was true. Furthermore, much better CVs were obtained when using FVIII-deficient plasma than when using buffer as prediluent, for all FVIII concentrates except for the intermediatepurity concentrate where the reverse was true, and sample B where CVs were equivalent. Overall, CVs were far worse than those obtained in controlled collaborative studies. Generally, however, CVs were better with chromogenic assays and predilution in FVIII-deficient plasma, as is recommended by the International Society on Thrombosis and Haemostasis/Scientific and Standardization Committee, particularly for higher purity and recombinant concentrates.