1.Invertebrate‐derived DNA (iDNA), in combination with high throughput sequencing, has been proposed as a cost‐efficient and powerful tool to survey vertebrate species. Previous studies, however, have only provided evidence that vertebrates can be detected using iDNA, but have not taken the next step of placing these detection events within a statistical framework that allows for robust biodiversity assessments. 2.Here, we compare concurrent iDNA and camera‐trap surveys. Leeches were repeatedly collected in close vicinity to 64 camera‐trap stations in Sabah, Malaysian Borneo. We analyse iDNA‐derived mammalian detection events in a modern occupancy model that accounts for imperfect detection and compare the results with those from occupancy models parameterized with camera‐trap‐derived detection events. We also combine leech‐iDNA and camera‐trap data in a single occupancy model. 3.We found consistent estimates of occupancy probabilities produced by our camera‐trap and leech datasets. This indicates that the metabarcoding of leech‐iDNA method provides reasonable estimates of occupancy and may be a suitable method for studying and monitoring mammal species in tropical rainforests. However, we also show that a more extensive collection of leeches would be needed to assess mammal biodiversity with a robustness similar to that of camera traps. As certain taxa were only detected in leeches, we see great potential in complementing camera‐trap studies with the iDNA approach, as long as the collection of leeches follows a robust and standardized sampling scheme. 4.Synthesis and applications. Here, we describe an approach to analyse detection records of mammals derived from leech samples using an occupancy framework that accounts for leech‐specific factors influencing the detection probability. We further combined camera‐trap and leech data, which lead to increased confidence in occupancy estimates. Our approach is not restricted to the processing of leech samples, but can be used for the analysis of other invertebrate DNA (iDNA) and environmental DNA (eDNA) data. Our study is the first step to shift the application of iDNA studies from opportunistic ad‐hoc collections to the systematic surveys required for long‐term management of wildlife populations.