Abstract <p>Determining reliable methods to effectively survey the abundance and species richness of wild bees is a prerequisite for an assessment of the status and trends of their populations. The present study evaluates the results of a three-year, three occasions per year, monitoring of wild bees conducted at nine localities across Vojvodina, a northern Serbian province, combining pan traps (PT) and transect walks (TW). The overlapping but partially distinct bee communities were captured via the two methods, with more species and specimens captured via TW than PT. The generalized linear mixed models were used to evaluate how total flower cover, dominant flower color and trap color affected the detected bee species richness and abundance. The TW results showed the effect of dominant flower color on bee species richness to be significant, while bee abundance was significantly influenced only by flower cover. PT results revealed that both species richness and abundance were significantly affected by the interaction between trap color and flower cover / dominant flower color. A comprehensive survey combining TW and PT (blue/white/yellow) is recommended in order to gain a more complete insight into the composition of bee species, taking into account the effects of floral and color context, especially in terms of their interaction.</p> Implications for insect conservation <p>The present study has highlighted the need to consider various aspects of wild bees monitoring for the sake of an adequate interpretation of results and the suitability of used methods, as a prerequisite for an efficient conservation of these key pollinators.</p>