Decreasing body size has been suggested as the third universal biological response to global warming after latitudinal/altitudinal range shifts and shifts in phenology. Size shifts in a community can be the composite result of intraspecific size shifts and of shifts between differently sized species. Metabolic explanations for the size shifts dominate in the literature but top down effects, i.e. intensified size-selective consumption at higher temperatures, have been proposed as alternative explanation. Therefore, we performed phytoplankton experiments with a factorial combination of warming and consumer type (protist feeding mainly on small algae vs. copepods mainly feeding on large algae). Natural phytoplankton was exposed to 3 (1(st) experiment) or 4 (2(nd) experiment) temperature levels and 3 (1(st) experiment: nano-, microzooplankton, copepods) or 2 (2(nd) experiment: microzooplankton, copepods) types of consumers. Size shifts of individual phytoplankton species and community mean size were analyzed. Both, mean cell size of most of the individual species and mean community cell size decreased with temperature under all grazing regimes. Grazing by copepods caused an additional reduction in cell size. Our results reject the hypothesis, that intensified size selective consumption at higher temperature would be the dominant explanation of decreasing body size. In this case, the size reduction would have taken place only in the copepod treatments but not in the treatments with protist grazing (nano- and microzooplankton).