Citizen Science Data (CSD) is increasingly contributing to the assessment of biodiversity and ecosystems. However, there is a need to evaluate the usefulness of CSD for different purposes. Ideally, CSD from populations should be evaluated against independent population data collected using a proper sampling design, but such data are lacking for almost all species. We propose an approach for evaluating CSD against environmental data. First, an evaluation model is formulated based on knowledge of how environmental variables affect population dynamics. Second, the hypotheses of the evaluation model are tested statistically. Support for the evaluation model is interpreted as support for the CSD. We applied the approach to six longhorn beetle species using Swedish data from 1930–2000. The evaluation model assumed that early summer temperature affects larval development time. We found support for the evaluation model in two species, and some evidence in its favour in one species. This suggests that the CSD from these species reflect true inter-annual variation. We also found statistical evidence for population trends in three to four species. In two of these, the evaluation model was supported thus providing particular support for the trend estimates. Lack of support for the evaluation model may be due to biological inaccuracy, the general characteristics of CSD, or low resolution of the environmental evaluation data. We also discuss alternative environmental data for evaluating CSD.