The Mastomys natalensis species complex, subdivided into genetically distinct species having diploid chromosome numbers 2n = 32 and 2n = 36, is a reservoir for several zoonoses including Lassa fever and plague. This report describes a study to determine whether these sibling species and three other rodent species have different potential as reservoirs for plague. It was found that M. natalensis (2n = 32) was significantly more resistant to experimental plague infection (50% survived inoculation with 120 000 Yersinia pseudotuberculosis subsp. pestis) than was M. coucha (2n = 36) (none of which survived doses of 190 Y. pseudotuberculosis subsp.pestis). In descending order of resistance were M. natalensis, Aethomys chrysophilus, M. coucha, Tatera leucogaster and A. namaquensis. No A. namaquensis survived inoculation of 10 or more plague bacilli.Previous reports on susceptibility to plague or other infections, which were based exclusively on findings in the universally distributed laboratory-bred Mastomys, are thus not necessarily applicable to the M. natalensis species as a whole but probably only to M. coucha. The Y. pseudotuberculosis subsp. pestis fraction-1 passive haemagglutination test appeared to be relatively insensitive in that only 5 out of 47 animals surviving experimental plague infection showed specific antibodies 6 weeks after challenge.The geographic distribution of human plague in southern Africa corresponds closely with that of the plague-susceptible species, M. coucha, while the resistant species, M. natalensis, predominates in areas where human plague has not been recorded. The role of A. namaquensis in the ecology of plague needs to be carefully studied and its possible importance in plague research should be investigated further.