Grazing is an established conservation tool for maintaining grassland habitats and under some circumstances may enrich arthropod assemblages. However, even if enrichment occurs, it is not granted that conservation value signified by rare and specialist species will also increase. To assess how some preset levels of grazing suit conservation aims, we studied spider assemblages of ungrazed, sparsely grazed and intensively grazed areas of a pasture in Hungary for three years by pitfall trapping and suction sampling. At ground level there was no significant difference among grazing areas, while at higher strata increasing grazing intensity negatively affected number of individuals and species. C-score analysis indicated equally neutral community assembly in all three grazing areas. All statistical methods that took into account species identity indicated virtually no difference between the spider assemblages of the sparsely grazed and ungrazed areas; however, there was a marked difference between these and the intensively grazed area. Spider species in the intensive grazing area had significantly lower affinity but wider tolerance for habitat naturalness, preferred more open habitats and had a lower rarity status. In the intensive grazing area a number of disturbance-tolerant species, among them agrobionts, were present, whereas the exclusion of rare or specialist species in the intensively grazed area occurred infrequently. The primary effect seen at the intensive grazing area was the opening of the spider assemblage to disturbance-tolerant species, while species richness was likely maintained by neighboring source populations. Overall, we experienced a marked decrease in the naturalness status of the spider assemblage in the intensive grazing area.