Ancient DNA has been extracted and sequenced from several animal and plant specimens. Previous considerations of the damage to ancient DNA have suggested that both the age and size of DNA fragments that can be retrieved and sequenced may be limited, the former to between several thousand and at most tens of thousands of years old, and the latter to at most a few hundred bases. A recent report of a 770 base pair (b.p.) sequence from the chloroplast gene rbcL from a Miocene Magnolia latahensis leaf indicates that both estimated limitations may be too conservative. Further work has indicated that analysis of Miocene fossil DNA can be replicated, and can, therefore, open up the prospects for future development of the field of molecular palaeontology. Successful amplification of fossil DNA is sometimes confounded by factors inherent to fossil DNA or to samples with minimal amounts of target DNA. Techniques that alter denaturation, reduce inhibitors and the problem of contaminants, and repair DNA prior to polymerase chain reaction amplification can increase the probability of success.