Abstract <p>Regions with exceptional bee diversity, such as South Africa, lack IUCN Red List assessments. This is particularly concerning, as 52% of South Africa’s bee species are endemic, yet nearly a quarter of the country’s natural habitat has been lost. However, Red List assessments of bees are challenging due to scarce data on species’ geographic ranges. Overcoming this requires methods that leverage available data to improve range estimates. We assess the extinction risk of specialist oil-collecting <i>Rediviva</i> Friese bees that have recently been re-surveyed and taxonomically revised. We estimate range sizes using both the IUCN-approach and species distribution modelling (SDM) and incorporate national spatial overlays to assess habitat loss, threats and protected area coverage. Using this integrative approach, we conducted the first quantitative Red List assessment of South African bees, refining upper-bound AOO estimates with SDMs and evaluating rates of habitat loss, protection, and vulnerability. We show that the IUCN- and SDM-approaches are best when combined with spatial overlays. Agriculture and forestry were identified as the main threats, contributing approximately 40% of habitat loss. While the IUCN-approach indicated limited protection, the SDM-approach captured up to 22% more potential habitat within protected areas. Although based on a subset of specialist bees, we provide a methodological framework broadly applicable to improving extinction risk assessments in other data-limited insect taxa.</p> Implications for insect conservation <p>Our integrative approach enhances the accuracy of Red List assessments for insects with limited occurrence records, supporting conservation prioritization and expanding taxonomic and geographic coverage of Red List assessments.</p>