Gaylor (1980) describes conclusions drawn from the initial analysis of the ED01 study. These conclusions failed to take adequate account of differences in time on study for animals at different doses. Several consequences of this are discussed. Two conclusions of Gaylor's are identified that are open to question. The ED01 liver and bladder tumor data, adjusted for time-on-study, are used to firmly reject several mathematical models that have been proposed for low-dose extrapolation. The complexities that arise from differences in time-on-study are shown to cast serious doubt on the general concept of low-dose extrapolation, when it is attempted independent of time to tumor. It is pointed out that the problems of statistical analysis that arise when dosed animals live longer than controls (as occurred in the ED01 study) raise questions about the design of lifetime feeding studies in general.