Pasteurella spp. sensu stricto and other taxa of Pasteurellaceae obtained from lesions in dogs or wounds inflicted by dogs represent a diagnostic enigma if based upon phenotypic characterization. Recently, taxon 16 was demonstrated to represent a new genus, 'Frederiksenia' improving separation of this taxon. However, other taxa, like biovar U24 of Biberstein, associated with respiratory – and urogenital tract infections in dogs and [Haemophilus] haemoglobinophilus/'H. canis' (Kilian 1976), have remained unclassified and unreported. Here we compare these taxa with taxon 35 of Bisgaard, a previously unreported group of Pasteurellaceae from dogs. A monophyletic group was formed based on partial rpoB and on 16S rRNA sequence comparison of strains of taxon 35, biovar U24 and [H.] haemoglobinophilus. Similarities within the group were 96.5 and 97.8% by rpoB and 16S rRNA, respectively, with Bisgaard taxon 5 (85.7%, rpoB) and [H.] felis (96.0%, 16S rRNA) as the closest related outside the group. Whole genome similarity estimated from recN similarity was 82% within the group and in accordance with other species of Pasteurellaceae while 57% to the closest related (Histophilus somni) was somewhat higher than the average level of 40%. The whole genomic sequences generated for strains representing [H.] haemoglobinophilus and taxon 35, respectively, resulted in a GGDC-index-2 of 0.00559 corresponding to 58% DNA reassociation, indicating closely related, but separate species. Phenotypic characterization showed that U24 and taxon 35 represent two biovars which differ in meso-inositol and D(+)galactose, while these groups differ from [H.] haemoglobinophilus in D(-)fructose and α-fucosidase. In conclusion, taxon 35, biovar U24 and [H.] haemoglobinophilus were shown to represent a monophyletic group. Phenotypic diversity, however, does not allow separation from the genera Pasteurella and 'Frederiksenia'. For the same reason Pasteurellaceae from lesions in dogs and wounds inflicted by dogs should always be subjected genotypic characterization to obtain a correct diagnosis.