Molecular detection systems used to analyse the gut contents of invertebrate predators have enhanced our understanding of trophic interactions, but do not distinguish between the methods of consumption. Many predators regularly scavenge, which could have profound implications for quantitative analyses of the dynamics of predation. We report the first quantified assessment of the potential error caused by scavenging in post-mortem measurements of predation in a slug = carabid system. An anti-slug monoclonal antibody was able to detect antigens from decayed slugs after surprisingly long periods, significantly longer on relatively sterile peat than on natural soil. On soil the half-life of antibody-detectable slug proteins was 8.2 days while on peat it was 11.5 days. When slugs that had decayed on soil for 100 h were fed to the carabid predator Pterostichus melanarius, slug proteins could still be identified after 6 h (but not 12 h) digestion. Fresh and decayed slug was eaten in equal quantities by the beetles suggesting no aversion to the latter. The results suggest that significant errors may be caused by scavenging leading to inaccurate interpretation of predation rates in the field.