The seasonal variation in the grazing effect of mixotrophic flagellates on bacterioplankton was assessed during an annual cycle in an oligotrophic coastal station in the northwest Mediterranean Sea. Ingestion rates of fluorescently labeled bacteria were estimated for different size categories of phytoflagellates (PF) and heterotrophic flagellates (HF) in short-term experiments and compared with long-term grazing estimates and published empirical models. The mixotrophic flagellates included haptophyte-like cells, cryptophytes, and dinoflagellates. The group-specific grazing rates (SGR) averaged 1.1 (3-5 mm PF), 1.3 (5-20 mm PF), 4.0 (,5 mm HF), and 15.4 bacteria individual 21 h 21 (5-20 mm HF). Lower SGR but higher abundances of PF resulted in an average mixotroph contribution of 50% to the total flagellate grazing. Remarkably, the effect was relatively high all through the year (35-65%). Regardless of the presence of chloroplasts, flagellates ,5 mm in size accounted, on average, for about 80% of total flagellate bacterivory and ingested a large percentage of their cell carbon per day from bacteria. Soluble reactive phosphorus concentration was negatively correlated with the ingestion rate of both groups of PF, suggesting that mixotrophic flagellates would be using their phagotrophic capability to obtain phosphorus when this nutrient is limiting. HF grazing activity showed a marked seasonality, with grazing being higher during the warmer seasons, and clearance rates were positively correlated with water temperature. Total bacterivory accounted for most of the bacterial production. Short-term and long-term bacterivory measurements were highly correlated, confirming that the smallest flagellates were the main causative agent of bacterial loss. The bacterivory values were also well correlated to a published empirical model that considers HF as the only bacterivorous. However, this model underestimated (up to 50%) total flagellate grazing during periods of high effect of mixotrophic flagellates.