A bioartificial liver (BAL) based on viable porcine hepatocytes can serve as a bridge to liver transplantation in patients with acute liver failure (ALF). To support liver functions, an adequate mass of hepatocytes is needed, which depends upon the cell density in the BAL device. This study evaluated the optimal density of hepatocytes within BAL devices that were constructed by perfusing porcine hepatocyte suspensions mixed with cytodex-3 into polysulfon hollow-fibers. The BAL devices were prepared with 6 different cell densities. The mass of hepatocytes in each device was evaluated for (a) cell viability, (b) ability to degrade diazepam, (c) ability to synthesize urea, (d) incorporation of [3H]-leucine into protein, (e) glucose-6-phosphatase activity, (f) total RNA content, and (g) p53 gene expression. Hepatocyte viability was about 90% in each device. With increasing hepatocyte density, the diazepam concentration in the medium decreased from 9.26 +/- 0.96 mg/L at 1 x 10(5) cells/ml to a minimum of 5.25 +/- 1.02 mg/L at 5 x 10(6) cells/ml and thereafter remained at low levels. Urea production and [3H]-leucine incorporation into protein increased progressively until the cell density reached 5 x 10(6)/ml and thereafter remained at high levels. Glucose-6-phosphatase activity and total RNA content stayed at high levels until the cell density reached 5 x 10(6)/ml and then progressively decreased. p53 gene expression differed from the other parameters, since it increased only when the cell density reached 5 x 10(7)/ml. In conclusion, the density of 5 x 10(6) cells/ml is a critical inflection point for most of the functional parameters, although p53 gene expression is not elevated at this cell density. These findings suggest that 5 x 10(6) cells/ml is the optimal hepatocyte density in the hollow-fiber BAL device.