A new automated kinetically determined fibrinogen assay was measured in plasmas of healthy subjects and three clinical cohorts (acute-phase reaction, liver cirrhosis and fibrinolytic therapy) that were expected to show normal, high and low levels of fibrinogen. The results were compared with the results of fibrinogen measurement using the derived method, the method according to Clauss and an immunological-nephelometric method. Altogether, the best correlation was achieved between the kinetic and the derived method. However, results from the derived method were generally higher than values obtained through the kinetic method. This was particularly true for high concentration levels above 400 mg/dl (patients with acute phase reaction) as well as for plasmas containing fibrin(ogen) degradation products and low concentrations of fibrinogen (below 150 mg/dl). Fibrinogen determinations in several commercial plasma pools with declared fibrinogen levels show remarkable heterogeneity when different methods were applied. To improve the discernment of fibrinogen determinations we suggest adjustment of standard preparations to international reference materials and the specification of the method used. Furthermore the attending physician is asked to cast a critical eye on fibrinogen values with regard to the used method of determination.