Abstract Since 1998, the Spanish Scientific Police has been analyzing both hypervariable segments (HVR1 and HVR2) of the mitochondrial control region by sequencing analysis. During this period, different sample types (e.g., blood, saliva, telogen hairs, hair shafts, nails, tissues, bone and dental remains) related to murders, sexual assaults, robberies and identification of human remains have been studied. Some problematic findings detected in routine forensic casework involving sequence heteroplasmy and one single nucleotide mismatch between forensic specimens and reference samples are described in this paper. In specimens which, in principle, due to apparent features should not give any problems when being analysed, a lack of amplification was detected within some of the subregions studied. Subsequently, it was proven that this was caused by point mutations in the primer binding sites. A substitution of the primers by non-conventional ones overcame the absence of reactions. Given that a proper statistical evaluation of the results using a large database is needed, samples from 120 individuals living in Spain have been analysed. The most frequent sequence found was 263-G; 315.1-C (differences compared to the reference sequence), being consistent with other Caucasian population studies.