The aim of this study was to compare the elimination of Bordetella pertussis clinical isolates, representing different genotypes in relation to alleles encoding virulence factors (MLST—multi-locus antigen sequence typing), MLVA type (multi-locus variable-number tandem repeat analysis) and PFGE group (pulsed-field gel electrophoresis) from the lungs of naive mice or mice were immunised with the commercial whole-cell pertussis vaccine, the acellular pertussis vaccine and the experimental whole-cell pertussis vaccine. Molecular data indicate that the resurgence of pertussis in populations with high vaccine coverage is associated with genomic adaptation of B. pertussis, to vaccine selection pressure. Pertactin-negative B. pertussis isolates were suspected to contribute to the reduced vaccine effectiveness. It was shown that one of the isolates used is PRN deficient. The mice were intranasally challenged with bacterial suspension containing approximately 5 × 10 7 CFU/ml B. pertussis. The immunogenicity of the tested vaccines against PT (pertussis toxin), PRN (pertactin), FHA (filamentous haemagglutinin) and FIM (fimbriae types 2 and 3) was examined. The commercial whole-cell and acellular pertussis vaccines induced an immunity effective at eliminating the genetically different B. pertussis isolates from the lungs. However, the elimination of the PRN-deficient isolate from the lungs of mice vaccinated with commercial vaccines was delayed as compared to the PRN ( +) isolate, suggesting phenotypic differences with the circulating isolates and vaccine strains. The most effective vaccine was the experimental vaccine with the composition identical to that of the strains used for infection.