AbstractCentral to theoretical studies of trophic interactions is the formulation of the consumer response to varying food availability. Response functions, however, are only rarely derived in mechanistic ways. As a consequence, the uncertainty in the functional representation of feeding remains large, as, e.g., evident from the ongoing debate on the usage of Ivlev, or Holling type I, II, and III functions in aquatic ecosystem models. Here, I refer to the work of Sjöberg in Ecol Model 10:215–225 (1980) who proposed to apply elements of the queuing theory developed in operational research to plankton–plankton interactions. Within this frame, food item processing is subdivided into two major stages which may operate with variable synchronicity. Asynchronous phasing of the two stages enhances the probability of long total processing times. This phenomenon is here termed feeding intermittency. Intermittency is assumed to determine the functional form of grazing kinetics, for which a novel grazing function containing a “shape” parameter is derived. Using this function, I evaluate the hypotheses that intermittency is influenced by (1) patchiness in the prey field (e.g., related to turbulence), and (2) the ratio of actual prey size to optimal prey size. Evidence for the first hypothesis arises from explaining reported variations in clearance rates of Acartia tonsa under different turbulence regimes. Further model applications to ingestion data for rotifers, copepods, and ciliates support the view that an increasing food size enhances intermittency and, this way, affects functional grazing responses. In the application to ciliate grazing, a possible prey density effect appears, possibly due to an intermittent activation of a feeding sub-stage. Queueing theory offers mechanistic explanations for transitions between Holling I-, II-, and Ivlev-type grazing. In doing so for variable prey size ratios, it may also refine size-based ecosystem models which are increasingly emerging in plankton ecology.