Until now the mechanisms of how recent historical land-use (hereafter called "habitat") changes in island ecosystems shape the distribution of individual insect species have been poorly understood in the field of conservation biology. In the present study we concentrate on the delicate equilibrium of the contribution of habitat island composition (i.e. habitat resource availability) with respect to island insect distribution patterns. In this context we study in detail the distribution patterns of four functional groups (herbivorous sucking insects, herbivorous chewing insects, spiders and other arthropod predators) of endemic, native and exotic arthropod species in a well-studied island of the Azores archipelago (Terceira). Within the bigger context of a standardized sampling program both for epigean and canopy insects and spiders we want to find out which species are truly rare and which are pseudo-rare species in each target habitat. Two dimensions of rarity were measured: abundance and habitat specialization. Two domains of rarity were identified: "among habitats" and "geographic". Some interesting patterns emerged. The high dispersal abilities of many insect and spider species together with the fact that many species from islands tend to be generalists imply that many species tend to be vagrants in several habitats and consequently are locally habitat pseudo-rarities. Two types of local pseudorare species were identified: "habitat (or land-use)" and "host plant" pseudorarities. Some species are rare in one habitat type whilst they are more common in another, often related habitat, or they are relatively rare in many habitats. This is a consequence of a "mass effect", with many species demonstrating a "source-sink" dynamics. Truly regionally rare species are those that are habitat specialists and many of them are threatened endemic species or recently introduced exotic species. We suggest several hypotheses for the patterns found, based on the former larger distribution and disturbance regimes of the native Laurel forest. Insufficient spatial replication in sampling can lead to the conclusion that numerous species appear to be rare because they were sampled in marginal sites or in the edge of their distribution. Since habitat occurrence is a less reliable predictor of the rarity status, more attention should be given to the standardized sampling of many habitats before extracting conclusions about the threatened status of a particular insect or spider species. Our results provide clear evidence that without adequate spatial data on abundance and habitat requirements, rarity status for insects and spiders on islands and elsewhere cannot be appropriately assessed.