A Bioartificial Liver (BAL) is not available to date. Human therapeutic applications of different systems of BAL as of 1998 are presented and discussed. It is concluded that--as for now--no artificial liver device has gained any importance for the treatment of liver failure, and that some critical issues in this field of research have not been sufficiently investigated and/or are not resolved. This review analyzes reports of clinical applications of BAL from the following research groups (numbers in brackets indicate patients treated with an artificial liver device): Demetriou/Rozga, Los Angeles (31 patients), Williams, London (1 patient); Gerlach, Berlin (1 patient); Strom, Virginia (5 patients). The BAL systems used in these studies cannot be directly compared because there are considerable differences in the quality and in quantity of the functional unit employed in bioreactors, the experimental design, patient selection, to mention just a few points. None of the systems investigated so far could convincingly prove its effectivity in replacing impaired liver function neither in animal models nor in a clinical application. It remains to be shown, whether liver cells cultured in bioreactors remain stable, i.e. viable and functionally active, for a sufficient period of treatment. Selected metabolic or detoxifying functions of the bioreactor are difficult to assess, since these functions do not necessarily serve as pars pro toto for the complex clinical presentation of liver failure and therefore cannot sufficiently validate any organ replacement system. Furthermore, since some applications combine biological units with other components, such as active charcoal, it be-comes even more difficult to assess the role of hepatocytes in these settings. Case reports of patients treated with BAL usually refer to BAL as a "successful bridge to transplant", thereby demonstrating the positive effect of orthotopic/auxiliary liver transplants in the treatment of acute liver failure rather than the potential benefit derived from an artificial liver device. Randomised studies have been proposed and urged for years in order to prove the effectivity of these systems which, in part, are already available for clinical use. Because of the heterogeneous patient group in question the design of such protocols will be a difficult task. It must be asked, whether currently used artificial liver systems have left basic science research too early; the use of "black box" applications in humans cannot draw its legitimation merely from the fact that an effective conservative treatment of liver failure is not available so far.