Plantations and secondary forests are becoming dominant components of many tropical forest landscapes. Yet we have an insufficient understanding of the value of these habitats for biodiversity conservation, and almost none for most arthropods in species-rich tropical forests. We sampled epigeic arachnids (Amblypygi, Araneae, Opiliones, Scorpiones, and Uropygi) in primary, secondary (14–19 years), and Eucalyptus plantation (4–5 years) forests in the Jari region of northeastern Brazilian Amazonia. We sampled five independent sites in each forest type between January and June 2005, collecting a total of 4824 individuals (3177 adults, 112 species), including 1864 adults (75 species) in Eucalyptus, 776 (60) in secondary forest, and 536 (72) in primary forest. We compared species richness, species-abundance distributions, and community structure, between the three forest types and identified the species that were characteristic of each forest type. Rarefaction analyses showed that undisturbed primary forest harbored significantly more species and a similar overall abundance as second-growth forest; while levels of species richness were similar between secondary forest and Eucalyptus. The species composition and abundance structure of arachnid assemblages was distinct in all three forest types. Considering all species sampled, 19% were only sampled in primary forest, 4% in secondary forest, and 19% in Eucalyptus. Most species sampled in plantation forests are known to be wide-ranging habitat generalists. Our data indicate that regenerating forests are not biological deserts (57% and 56% of species sampled in primary forest were also captured in secondary and plantation forests respectively) and can, therefore, help mitigate some of the negative effects of deforestation for epigeic arachnids. However, these replacement habitats do not provide a substitute for primary forest and may fail to conserve many of those species most at risk from extinction.