Invasive common wasps (Vespula vulgaris L.) are predators of invertebrates in Nothofagus forests of New Zealand. We reduced wasp densities by poisoning in three sites over three y. We predicted an increase in the number of invertebrates and a change in the community composition in sites where wasps were poisoned (wasps removed) relative to nearby sites where wasps were not poisoned (wasps maintained). Wasp densities were significantly reduced by an average of 58.9% by poisoning. Despite this reduction in wasp densities, native bush ants (Prolasius advenus Forel) were the only taxa that was significantly influenced by wasp removal. However, contrary to our predictions there were more ants caught in pitfall traps where wasps were maintained. We believe that the higher abundance of these ants is probably because of the scarcity of honeydew in wasp-maintained sites and compensatory foraging by ants in these areas. Otherwise, our results indicated no significant effects of reduced wasp densities on the total number of invertebrates, or the number of invertebrate families, observed in pitfall or Malaise traps. An analysis of community composition (permutational multivariate analysis of variance) also indicated no significant difference between wasp-removed or wasp-maintained communities. The most parsimonious explanation for our results is that although we significantly reduced wasp numbers, we may not have reduced numbers sufficiently or for a sufficiently long period, to see a change or recovery in the community.