The relationship between toxicokinetics and time-dependent PAH toxicity to Hyalella azteca was examined to test the constant critical body residue (CBR) model. A constant CBR model is based on the assumption that the body residue for 50% mortality is constant for each PAH across exposure times. With a constant CBR, kinetic parameters determined through kinetic experiments would be similar to those estimated from time series toxicity data. Time-dependent toxicity was investigated using three types of data:  time series LC50 data, LT50(c), and CBR values measured at multiple exposure times for live and dead animals. Kinetic parameters were measured independently. The constant CBR model did not predict the PAH toxicity time course for H. azteca. Since a first-order kinetic model predicted the bioaccumulation of the parent PAH except for naphthalene, this result is not due to a failure to predict the internal dose (body residue). The influence of metabolites on toxicity was negligible except for naphthalene. The LC50 values at multiple exposure times decreased to an incipient lethal concentration after H. azteca reached steady state. Measured CBR values also decreased with increasing exposure time. Thus, the time course of PAH toxicity is determined not only by the bioconcentration kinetics but also by the cumulative toxicity with increasing exposure time. Therefore, time-to-death or hazard models must be developed as a complement to toxicokinetic models to describe and predict the toxicity time course.