answer:“There is one problem that DNA studies of population movements within the past 10,000 years do not have: dating of the samples, as radiocarbon dating is considered to be accurate up to 50,000 years. Beyond that, other methods are needed (Hebsgaard et al, 2005), which increases the margin for error. Nevertheless, most researchers in the field accept that the oldest DNA sample sequenced so far is from plant material around 400,000 years old (Willerslev et al, 2003). From animals, sequences at least 70,000 years old have been obtained—Gilbert and others published a sequence of bison DNA that was at least 64,800 years old (Gilbert et al, 2004).” This is from Ancient DNA Research Goes Nuclear, an article by Philip Hunter, © 2006, published in the European Molecular Biology Organization. It’s on the NCBI (the National Center for Biotechnology Information) web site, which is in conjunction with the National Institute of Health, and their journal database, PubMed is one of the most used in the country. As far as peer-reviewed scientific sources go, this is golden. The next paragraph is also very interesting, and sheds further light on your question: “This begs the question of what is the theoretical age limit for recoverable DNA. Various attempts have been made to calculate the theoretical survival rate for DNA at different temperatures, but as Gilbert pointed out, these have been based on simple degradation models that are known to be fairly inaccurate. The generally accepted limit is around 1 million years, certainly no more than 2 million, which makes it impossible to recover, for instance, DNA from dinosaurs. Yet DNA damage occurs in many ways, so it is possible that the absolute limit may be considerably greater, even if a sequence of that age has never been recovered.” Wow, isn’t this stuff incredible?? :)