This study examined maximum life span predictions obtained with the Gompertz mortality rate model, which assumes that there is a constant rate of acceleration in the age-related mortality of adult populations. The influence of population size N on the maximum life span (tmax) was shown to be small, because the numeric impact of N is reduced to ln[ln(N)]. In contrast, the Gompertz exponential mortality coefficient alpha has much more influence on the tmax, which varies as 1/alpha. Examination of select mammals and birds showed that tmax as reported for local populations agrees very well with that calculated from mortality rate coefficients for these local populations. However, the tmax as reported from the world literature, which is designated here as the "world record, " shows major discrepancies for some species from the predicted tmax based on the local population. We demonstrate that these discrepancies are not due to population size, but represent other factors that may include genotype, diet, and environmental dangers. Potential increases in human tmax will depend mostly on slowing the age-related acceleration of mortality. If the degree of mortality rate slowing achieved in rats by diet restriction is applied to humans, then the median human life expectancy would approach the present tmax of 120 years.