Invasive species are a global problem but most studies have focused on their direct rather than indirect ecological effects. We studied litter and soil-inhabiting invertebrate communities on 18 islands off northern New Zealand, to better understand the indirect ecological consequences of rat (Rattus) invasion. Nine islands host high densities of burrowing procellariid seabirds that transport large amounts of nutrients from the ocean to the land. The other nine have been invaded over the past 50–150 years by rat species that have severely reduced the density of seabirds by preying on eggs and chicks. Invaded islands had lower densities of seabird burrows but deeper forest litter than did the uninvaded islands, indicative of rats reducing disturbance effects of seabirds. However, despite deeper litter on the invaded islands, eight of the 19 orders of invertebrates that we measured were significantly less abundant on invaded islands. Furthermore, three soil-inhabiting micro-invertebrate groups that we measured were significantly less abundant on invaded islands. These differences probably result from rats thwarting transfer of resources by seabirds from the ocean to the land. We also investigated community-level properties of each of three test groups of invertebrates (minute land snails, spiders and soil nematodes) to illustrate this process. Spiders were equally abundant on both groups of islands, but showed lower species richness on the invaded islands. The other two groups showed no difference in species richness with island invasion status, but were more abundant on uninvaded islands. Reduced abundance of soil nematodes on invaded islands provides strong evidence of indirect consequences of seabird reduction by rats, because nematodes are unavailable to rats as prey. We predict that if rats are eradicated from islands, components of below-ground invertebrate dependent on seabird-mediated soil conditions may take considerable time to recover because they require subsequent seabird recolonisation.