Ecological citizen science initiatives are growing in popularity with the increasing realisation of the potential for occurrence records to contribute information on biodiversity. Citizen science data are, however, justifiably criticised for misidentification, uneven sampling, incomplete detection or selective reporting. Here, we test the accuracy of citizen science data for UK social wasp (Vespinae) species’ distributions. We compared data collected over 2 weeks by members of the public setting out baited traps across the UK and sending captured specimens for expert identification [1294 locations; 6680 wasps; three dominant species Vespula vulgaris (44%), Vespula germanica (44%) and Vespa crabro (6%)], with a four‐decade long‐term data set established by the Bees, Wasps and Ants Recording Society (BWARS). The citizen science data were significantly less spatially biased than the long‐term data, but they were more urban‐biased. Species distribution modelling showed that, for Vespa crabro, just 2 weeks of citizen science generated coverage comparable to more than four decades of expert recording. Overall, we show that citizen science can be an extremely powerful and robust method for mapping insect diversity and distributions. We suggest that cautious combination of citizen science data with long‐term expert surveying could be a highly reliable method for monitoring biodiversity.