Global studies of within-group genetic variation have revealed a tendency for some traits, but not all, to show higher heterozygosity in sub-Saharan African populations. Although excess African diversity has been interpreted as reflecting a greater "age" of sub-Saharan African populations, more recent research has shown that this excess is more likely a consequence of a larger African long-term effective population size. The observation that certain traits, particularly classic genetic markers and RFLPs, do not show this pattern has been interpreted as ascertainment bias. Here, I examine another possible factor: that excess African heterozygosity is in part a function of mutation rate. Simple equilibrium and nonequilibrium models of absolute excess heterozygosity are examined. The results indicate that there is little excess African heterozygosity for traits with low mutation rates and greater excess heterozygosity for traits with moderate to high aggregate mutation rates. Observed data are consistent with these models. Also, depending on population size and time depth, traits with high levels of mutation might show less excess heterozygosity than those with moderate to high mutation rates. Another measure of diversity, mean sequence divergence, shows an increase in excess diversity for traits with high mutation rates.