Mycoplasmas which had in various ways proved to be related to Mycoplasma mycoides were studied with a view to classification and nomenclature. Representative strains of M. mycoides — subspecies mycoides (PG1) and subspecies capri (PG3) — and of the as yet unclassified bovine serogroups 7 (PG50) and L (B144P) were compared serologically and by DNA-DNA hybridization. The serological methods used (metabolism inhibition, growth inhibition and immunofluorescence) are usually considered to be species-specific. Conclusions drawn on the basis of the premises just mentioned give rise to a dilemma, because three different situations exist. In (1), significant serological differences were disclosed, but the DNA homology was great. In (2), the serogroups 7 and L showed conformity between hybridization and serological findings, whereas, in (3), hybridization showed a distinct separation, but a serological cross-reaction does occur. If the results of hybridization are regarded as decisive, the two subspecies of M. mycoides can be preserved, but this is at variance with the general perception that the serological methods used here are species-specific. If, on the other hand, the serological results are chosen to be decisive, PG1 and PG3 must be regarded as two separate species, and consequently Group L/7 may also be a separate species. As regards the relation between group L/7 and M. bovigenitalium, the results of hybridization clearly show that they are two different species. As a consequence of this, it must be accepted that cross-reactions may occur — also in growth inhibition — between different species which, genetically, are even far removed from each other. On account of the fundamental nature of these problems, the authors prefer to leave the decisions — also as to the nomenclature — to the Subcommittee on the Taxonomy of Mycoplasmatales.