Genotypic testing includes several steps (RNA purification, RT-PCR amplification, DNA sequencing, sequence editing and analysis) that should be individually controlled. In our laboratory, we have added to this step-by-step internal control a final phylogenetic quality control: this is performed every time a sequence is obtained from a patient previously subjected to the same test. Each sequence with this characteristic is routinely compared with sequences from previous samples of the same patient by multiple alignment and a neighbor-joining tree by using Kimura two-parameter method is constructed. To validate the quality control procedure, we have aligned and calculated the mean similarity of the reverse transcriptase (first 984 nucleotides) and protease (whole gene) sequences from 30 patients whose virus was completely wild-type for both reverse transcriptase and protease. In the same tree, we have added the sequences obtained from 5 out of the 30 patients, tested at a second time point. The wild type sequences have shown a mean inter-sample divergence of 2.9%, and all the sequence pairs from individual patients clustered together in the tree constructed with the nucleotide sequences, while the tree constructed with the inferred aminoacid sequences did not always permit to cluster the sequences from the same patients. This indicates that: 1) the phylogenetic analysis of nucleic acid sequences can be useful to rule out sample mix-up; 2) the belonging of a sequence to each individual patient can efficiently be assessed also in the cases of extreme divergence in terms of drug resistance mutations.