Urban environments can serve as important wildlife habitats, especially for insects. This study investigates how (peri-)urban structure, habitat characteristics, and management influence butterfly populations in Leipzig, Germany. To inform better planning and management, the citizen science project VielFalterGarten works in collaboration with the city council and an NGO, inviting citizens to record butterflies as indicators of urban biodiversity using structured 15-minute counts with repeated visits at registered sites. We analysed count data with validated observations of over 6000 individuals of 41 species. We found that landscape-level parameters within a 500-metre radius, such as presence of shrubland and low shares of built-up areas were positively associated with butterfly abundance. Local management like the presence of flower strips supported abundance, reduced mowing regime increased species diversity. Our results demonstrate species’ traits acting as mediators, especially voltinism, wing size, and diet specialism. Urban greenspaces with higher shares of agriculture and less built-up areas, reduced mowing intensity or flower strips supported smaller, univoltine, and more specialist species. Analyses for five common species indicated high degrees of heterogeneity in the factors explaining local abundance. Our results indicate larger-scale negative effects of intense management of urban parks dominating over local-scale conservation measures. We therefore recommend conservation and restoration of low-intensity habitats including flower strips, shrubs, and reducing mowing frequency and intensity rather than only planting trees as required by the Nature Restoration Regulation. These findings underscore the complexity of urban butterfly ecology and highlight the need for nuanced approaches to conservation strategies. Our study demonstrates the value of systematic 15-minute counts as a scalable and accessible citizen science monitoring tool, particularly suited to assess small-scale urban biodiversity patterns and management impacts with non-expert participation. Our findings reaffirm the value of citizen science for informing urban management and for involving citizens in urban landscape planning and conservation.