The P function of vesicular stomatitis virus (VSV) is defined as the viral function which results in a reduced rate of total protein synthesis (viral plus cellular) arising from a nonspecific reduction in the efficiency of the translational machinery in infected cells. The existence of P function has been challenged by Lodish and Porter who were unable to detect it in L-strain mouse cells infected with wild-type VSV (HR) or, as expected, with the P- mutant, T1026-R1. Although other groups have subsequently confirmed the existence of P function and the difference between HR and T1026-R1, we have sought an explanation for the difference between Lodish and Porter's results and those of other laboratories. We show that the VSV P function depends on the phase of the growth cycle of infected L-cell cultures. In very early exponential phase, as used by Lodish and Porter, HR has very little demonstrable P function; as the growth cycle proceeds toward stationary phase, P function becomes more and more manifest. Under the same conditions, T1026-R1 shows no P function throughout the growth cycle. Furthermore we show that the VSV M protein mutant tsG31 has a P++ phenotype reducing total protein synthesis below that seen with wild-type HR. P function can be observed in cells infected with tsG31, even early in the exponential phase of the cellular growth cycle.