Genetically modified organisms (GMOs), which produce Bacillus thuringiensis (Bt) toxins, are widely used in agriculture in some parts of the world. Despite this, ecotoxicological methods, tailored to GMOs, are lacking to assess effects on aquatic environments. With the objective to investigate a food-related exposure pathway for aquatic shredders, we used a new food-spiking method while caddisfly larvae (Chaetopteryx spec., Sericostoma spec.) served as test species. Pure Cry1Ab toxins were spiked on black alder leaf discs and subsequently used in a feeding experiment. The toxin did not influence larval mortality compared to the control. The results, however, showed significant effects on larval lipid content (Chaetopteryx spec.) and development (Sericostoma spec.) at concentrations of 17.2 and 132.4 ng Cry1Ab/mg leaf, respectively. These changes are indicative for impacts on the fitness of the specimen and thus relevant in a risk assessment context. Ultimately, the food-spiking method allowed applying different Bt toxin concentrations leading to the establishment of dose-response relationships for various response variables. The use of long test durations and sublethal endpoints (consumption, lipid content, growth, larval instars) is, moreover, advisable when testing GMO effects.