C/GT/A nucleotide alterations have been shown to hamper the straightforward interpretation of mitochondrial DNA sequence data derived from ancient tissues. Attempting to characterise this finding with respect to nuclear DNA, we contrasted two established protocols: (i) an enzymatic repair of damaged DNA, thereby translating and closing nicks in the DNA, and (ii) the application of N-phenacylthiazolium bromide, which cleaves glucose-derived protein crosslinks, presumably derived from Maillard reactions. We used medieval human bones that were refractory to standard PCR procedures. Due to negligible presence of short tandem repeat loci and also mitochondrial sequences, the extracted ancient DNA needed a higher copy PCR system to yield amplification products. The chosen PCR target was specific alphoid repetitive DNA with an experimentally determined minimum of 1000 copies per haploid genome. Alphoid repeat segments were generated from both contemporary DNA and DNA extracts of two human skeletons dating from 450-600 AD (omitting uracil N-glycosylase pre-treatment of the extracted samples), and were subsequently cloned and sequenced. The sequences were evaluated for the number and type of nucleotide alterations noted after the different pre-treatments, and were compared to our alphoid consensus sequence generated from modern DNA. Both methods failed to reflect the expected 32% variability among single alphoid repeats (accounting for locus-specific differences and polymerase errors) as well as to display the actual 2.88 ratio of transitions to transversions. Our data obtained from high-copy-number nuclear DNA mirror the phenomenon of sequence deviations observed in mitochondrial DNA extracted from old specimens.