Resistance variability was evaluated for five rootstock: three Myrobalan plum (Prunus cerasifera Ehr.) genotypes (P.1079, P.2175, and P.2032) grown from in vitro plantlets, one peach (P. persica (L.) Batsch `GF 305') grown from seeds, and one peach-almond hybrid (P. persica × P. amygdalus Batsch `GF 557') grown from rooted cuttings. Twenty-two root-knot nematode populations from different origins were used: Meloidogyne arenaria (Neal) Chitwood (six populations), M. incognita (Kofoid and White) Chitwood (eight populations), M.javanica (Treub) (four populations), M. hispanica Hirschmann (one population), M. hapla Chitwood (two populations), and an unclassified root-knot species (one population). The study was conducted under greenhouse conditions for 1 and 2 months. No galling or nematode reproduction was observed in P.1079 and P.2175, which should be considered immune; P.2032 showed the highest galling and nematode counts when inoculated with M. hispanica and M. javanica. In P.2032, a high proportion of males was recovered in populations that had a limited development. Because the populations of the first four Meloidogyne species reproduce by obligatory mitotic parthenogenesis, high sex ratio maybe the expression of a late form of resistance. Host suitability of `GF 305' was highly variable among M. arenaria and M. incognita populations. A lower relative variation was observed in M. javanica. `GF 557' was resistant to M. arenaria and M. incognita except for one population of M. arenaria that was weakly aggressive and susceptible to M. javanica. Consequently, resistances specific to the genus Meloidogyne for the Myrobalan plum genotypes P.1079 and P.2175, specific to the nematode species for `GF 557', and specific to the nematode population for `GF 305', were evidenced. This study indicates that, in rootstock selection procedures, it is important to test resistance to several populations within the same nematode species.