For a genetic analysis of ancient human populations to be useful, it must be demonstrated that the DNA samples under investigation represent a single human population. Toward that end, we have analyzed human DNA from the Windover site (7000-8000 BP). MHC-I analysis, using allele-specific oligonucleotide hybridization to PCR amplified Windover DNA, microsatellite analysis by PCR of the APO-A2 repeat and mtD-loop 3' region sequencing on multiple individuals spanning nearly the full range of estimated burial dates all confirm the hypothesis that there is a persistence of both nuclear and mitochondrial haplotypes at Windover throughout its entire period of use. Thus, Windover can be considered a single population. Neighbor-joining tree analysis of mtDNA sequences suggests that some mitochondrial types are clearly related to extant Amerind types, whereas others, more distantly related, may reflect genetically distinct origins. A more complete sequence analysis will be required to firmly resolve this issue. Calibrating genetic relationships deduced by tree analysis, radiocarbon dates and burial position, yields a human mtD-loop DNA rate of evolution of 3700 to 14,000 years per percent change. Both values are within the range of recent, independently calculated values using estimates of evolutionary divergence or theoretical population genetics. Thus we are beginning to realize the promise of ancient DNA analysis to experimentally answer heretofore unapproachable questions regarding human prehistory and genetic change.