Size-selective grazing on bacterioplankton by phagotrophic nanoflagellates was analyzed and modelled. The proposed model resembles a Monod equation (Michaelis-Menten-like expression) in which clearance rates by heterotrophic nanoflagellates depend on bacterioplanliton biovolume. Larger bacteria were ingested faster than smaller bacteria. The proposed model was in agreement with experimental data from different authors, both from cultures and natural assemblages, Size-selective grazing efficiency was analyzed as the ability of phagotrophic nanoflagellates to discriminate between cells differing in volume by a factor of two, e.g., corresponding to dividing and nondividing bacteria. Unlike previously published models, the proposed model suggests that phagotrophic nanoflagellates could be highly effective size-selective grazers for small bacteria (<0.1 mu m(3)), which are the most common bacterial sizes in planktonic systems. However, phagotrophic nanoflagellates were unable to size-discriminate dividing and nondividing bacteria for volumes >0.1 mu m(3). These results strongly support the hypothesis that heterotrophic nanoflagellate grazing, by discriminating between different sizes of dividing and nondividing bacteria, may actually be regulating bacterial size and growth rate in natural aquatic ecosystems.