Almost every animal trait is strongly associated with parasitic infection or the potential exposure to parasites. Despite this importance, one of the greatest challenges that researchers still face is to accurately determine the status and severity of the endoparasitic infection without killing and dissecting the host. Thus, the precise detection of infection with minimal handling of the individual will improve experimental designs in live animal research. Here, we quantified extracellular DNA from two species of endoparasitic worm that grow within the host body cavity, hairworms (phylum Nematomorpha) and mermithids (phylum Nematoda), from the frass of their insect host, a cave wētā (Orthoptera: Rhaphidophoridae) and an earwig (Dermaptera: Forficulidae), respectively. Frass collection was done at two successive time periods, to test if parasitic growth correlated with relative DNA quantity in the frass. We developed and optimised two highly specific TaqMan assays, one for each parasite-specific DNA amplification. We were able to detect infection prevalence with 100 % accuracy in individuals identified as infected through post-study dissections. An additional infection in earwigs was detected with the TaqMan assay alone, likely because some worms were either too small or degraded to observe during dissection. No difference in DNA quantity was detected between sampling periods, although future protocols could be refined to support such a trend. This study demonstrates that a non-invasive and minimally stressful method can be used to detect endoparasitic infection with greater accuracy than dissection alone, helping improve protocols for live animal studies.