The detection of extinctions at local and regional scales in many biodiversity hotspots is often hindered by the lack of long-term monitoring data, and thus relies on time series of occurrence data. Citizen science has repeatedly shown its value in documenting species occurrences, mostly in very recent years. This study investigates the effectiveness of untargeted citizen science records in discarding the possibility of local extinctions in butterfly populations across all Italian National Parks. We addressed three research questions: i) the ability of citizen science data to supplement existing knowledge to complete occurrences time series, ii) the impact of functional traits determining species appearance on data collection, and iii) the interplay between participant engagement and species appearance in the amount of diversity recorded on the iNaturalist platform. Our analysis of 47,356 records (39,929 from literature and 7,427 from iNaturalist) shows that the addition of iNaturalist data fills many recent gaps in occurrence time series, thus reducing the likelihood of potential local extinctions. User effort strongly interacts with species size, distribution, and length of flight periods in determining the frequency of records for individual species. Notably, records from more engaged users encompass a higher fraction of local biodiversity and are more likely to discard local extinctions, and these users are less affected by species size. We also provide updated butterfly checklists for all Italian National Parks and a new R package to calculate potential extinction over time. These results offer guidance for protected areas, conservationists, policymakers, and citizen scientists to optimise monitoring of local populations.