Changes in bee fauna, such as the disappearance of certain species or increasing abundance of others, are very important. The common belief that pollinating insects are facing problems also prompts detailed studies of the bee fauna in order to track on-going changes. Assessment of the state of bee communities in particular ecosystems or ecosystem complexes within a landscape, tracing the course of trends in fauna and also predicting their future structures resulting from current changes, are only possible if sampling is carried out at an appropriate frequency, so that representative materials are obtained. The aim of the present study was to determine what sampling intensity during the growing season would enable the collection of representative materials to evaluate species diversity of bees in a study area. Repeated bee sampling at monthly intervals throughout the period of activity of bees resulted in the identification of 73 bee species, corresponding to 51.4 % of the estimated number of bee species in the study area. When samples were obtained twice a month, 93 bee species were captured, accounting for 65.5 % of the estimated number of bee species. When sampling took place nearly four times a month, 108 bee species were captured, making up 76 % of the estimated number of bee species; the materials obtained at this sampling rate may be regarded as representative. The importance of accounting for the phenology of a given animal group during a sampling effort is also emphasised.