Stable isotope enrichment experiments offer a potentially powerful way to examine the base of aquatic food webs, but interpretation of the data from these experiments may be confounded by problems such as selective ingestion/assimilation of bulk food sources by consumers, variable tracer enrichment over time, and the failure of consumers to approach isotopic equilibrium with the tracer in their diets over the course of the experiment. Our study examined data from a stable isotope addition experiment in which ¹⁵N-labeled NH₄ was added to a midwestern US stream. A compartment model was used to provide insights into the kinetics of ¹⁵N uptake and release from algae, heterotrophic microbes colonizing detritus, and invertebrate consumers. The model accounted for temporal variation in the degree of isotopic enrichment and did not require the assumption of isotopic equilibrium between consumers and their diets. The importance of instream production (i.e., growth of algae and microbes within the study reach during the experiment) relative to allochthonous and upstream inputs was ∼38 to 50% for heptageniids and Psephenus, 10 to 20% for Orconectes propinquus, Gammarus, hydropsychids, and larval Stenelmis, and <10% for the filter-feeding Similium and the unionid mussel Pleurobema sintoxia. The alternative choices of algae or heterotrophic microbes as the basis of consumer diets made little difference in these estimates, even though the microbes became more ¹⁵N-enriched than the algae because microbes had higher turnover rates. These results were subject to a number of caveats, and guidelines for experimental design are suggested for future studies to help address some of these problems.