The "multiple-indicator dilution method" of measuring hepatic transport kinetics is subjected to simulation analysis. The objective is to examine the errors that may arise from treating the nonexchanging vasculature as a simple delay and to study the information content of simulated venous outflow curves. We find that the method cannot be counted on to provide consistently reliable estimates of either the transport rate constants or the sinusoidal volume. Estimates of the rate constant for irreversible solute removal from within liver cells are especially likely to be wrong. We suggest an alternative formulation of the governing differential equations that can substantially improve the estimates of the uptake parameter. These estimates may otherwise be subject to large systematic errors. Finally we discuss why the steady-state extraction fraction computed from the fitted parameters should be checked against a model-independent estimate obtained directly from the areas under the outflow curves. A method for making this comparison is at hand and should prove useful as a minimum criterion of internal consistency in animal experiments.