Agri-environment schemes (AES) provide an important mechanism for environmental improvement with the potential to benefit many taxa, yet evidence of national scale benefits is mixed. Multiple sources of evidence are available to assess AES effects, with different strengths and weaknesses, but most existing studies use a single dataset to evaluate AES impacts. We developed an approach analysing multiple datasets to assess relationships between AES and species abundance, richness and diversity, using the example of butterflies in England. We analysed data from a study specifically designed to assess AES effects (the LandSpAES study) alongside two different citizen science UK Butterfly Monitoring Scheme (UKBMS) surveys. UKBMS surveys were not designed to evaluate AES effects, but they provide better spatial coverage across the agricultural landscapes of England. We compared AES relationships between the three datasets, using a generalised AES gradient method to allow integration of different AES options, including the creation of habitat features such as wildflower strips and the restoration of semi-natural habitats. We assessed AES effects at both local (1 km) and landscape (3 km) scales. We found that AES in the surrounding landscape was positively associated with butterfly community responses in all three datasets and some evidence that local-scale AES was positively associated with butterfly richness. The smaller size of the LandSpAES study led to wider confidence bounds around effect sizes, but the careful design provided assurance that potentially confounding effects were accounted for. The wider spatial coverage of the citizen science datasets increased confidence that results can be extrapolated to the national scale. Synthesis and applications. Our results provide support for positive effects of AES on butterflies in England from multiple sources of evidence, providing confidence that these schemes are providing tangible benefits for butterflies. Our recommendations for managers and policy makers are (1) multiple data sources should be considered for AES monitoring and evaluation, exploiting the strengths of different data types; (2) AES intervention over larger spatial areas than individual fields and farms should be considered when aiming to provide benefits for butterflies.