Grazing on planktonic bacteria by microzooplankton was estimated by separating bacteria from the larger plankton with 1μm pore Nuclepore filtration and measuring changes in bacteria in filtered and unfiltered samples over 24 hours. In the absence of grazers, bacteria increased linearly. The regression coefficient of linear increase was used to estimatein situ bacterial production. When grazers were present, the changes in bacteria concentration usually took the form of a linear decline, and grazing was estimated by subtracting the regression coefficient of the unfiltered sample from that of the 1μm filtrate. Results from the Essex estuary-coastal system of northern Massachusetts show grazing and production at rates that indicate a daily turnover of the standing crop of bacteria, with highest values in mid-estuarine waters. Experiments on the size distribution of grazing showed that microzooplankton from 1–3μm were responsible for most of the observed decrease in bacteria. It was suggested that the basic pattern of linear increase of the bacteria in the absence of grazing reflects density-dependent limitation by substrate present at the outset of the incubation and is indicative of a population that has been maintained around the mid-point of the logistic growth curve by grazing.