Although survival analysis is a well-established mathematical discipline, there seem to be almost no attempts in survival modeling for experimentally virus-infected laboratory animals. We have taken up a stochastic approach originally developed by Shortley in the sixties and have applied it to three different types of experimental data: to virus titer determination, to the dose dependence of the mean survival time and to single survival curves. Experience concerning parameter estimation is reported and new ways of working with the model parameters are proposed. A standard mean survival time is defined and suggested as a new quantitative measure of virulence. Moreover, for the comparison of two experiments for which the amount of virions inoculated is kept fixed, but for which other parameters may vary, a new scheme of systematizing survival data from experimentally virus-infected laboratory animals is proposed. It is very likely that the model can be also applied to cancer survival data or any other infectious pathogen.