The overall persistence, Pov, has been suggested as an indicator that integrates single media half-lives and phase partitioning. However, the application of this indicator is hindered by the fact that there are different definitions of Pov and that the most common measure of Pov, the residence time at steady state, depends heavily on the release pattern. Here, different measures are compared that can be used to compute Pov from the mass of a chemical in the environment as a function of time in a dynamic multimedia model. These measures include the equivalence width, which corresponds to the residence time at steady state, the mean value of the function of mass vs time ("mean time"), and the clearance time. The temporal remote state is defined as the state of the model system long after the stop of emissions; it is independent of the release pattern of a chemical. The mean time in the temporal remote state reflects the long-term removal rate of a chemical and is used as a persistence benchmark. Correlations of the three measures of Pov to the persistence benchmark are analyzed. The maximum equivalence width, which is obtained with release to one of the media air, water, or soil, is a good approximation of the mean time in the temporal remote state. Because the maximum equivalence width is identical to the maximum residence time at steady state, the mean time in the temporal remote state can be estimated with existing steady-state models.