The serological and parasitological tests used for Trypanosoma brucei gambiense human African trypanosomiasis (HAT) diagnosis have low specificity and sensitivity, respectively, and in the field, control program teams are faced with subjects with positive serology but negative parasitology who remain untreated. The aim of this work was to explore, using PCR tool, the significance of these aparasitemic serological suspects. Since discordant PCR results have been observed earlier with different extraction methods, two DNA extraction methods were compared (the Chelex 100 resin and the DNeasy Tissue kit). The study was conducted on 604 blood samples: 574 from parasitologically confirmed patients, aparasitemic serological suspects and endemic controls collected in Côte d'Ivoire and 30 from healthy volunteers collected in France. No significant differences were observed between the PCR results obtained with the two extraction methods. Concerning PCR, problems of reproducibility and discordances with both serological and parasitological test results were observed, mainly for the aparasitemic serological suspects. In addition to previous results that pointed to the existence of non-virulent or non-pathogenic trypanosome strains and of individual susceptibility leading to long term seropositivity without detectable parasitaemia but positive PCR, the results of this study support the notion of a long lasting human reservoir that may contribute to the maintenance or periodic resurgences of HAT in endemic foci.