ABSTRACT: Bottle incubation experiments are widely used in mesozooplankton grazing studies. However, we have shown here that traditional particle removal experiments with Calanus finmarchicus and C. helgolandicus as grazers on naturalplankton may yield low or even statistically significant (p < 0.05) negative grazing estimates, even though negative grazing rates are impossible. Low grazing rates are often reported, especially on smaller prey types, despite abundant food andsignificant egg production. Microzooplankton, such as ciliates, show higher biomass-specific grazing rates on algae than do copepods and other mesozooplankton. Instead, copepods often selectively feed on the microzooplankton. Thus, apparent negative rateswould be expected when the release of microzooplankton grazing pressure outweighs the copepod grazing rates on the same food items in the incubation bottle. We show that this potentially large bias increases with microzooplankton community grazingpressure in the control. A simplified general method to correct for this bias is presented and compared with the original method (Nejstgaard et al. 1997, Mar Ecol Prog Ser 147:197-217). Although complexity and the need for taxonomic accuracy are reducedin the general method, the results are not significantly different between the 2 methods. Both methods also show a good fit with ingestion rates estimated from faecal pellet production. We suggest that the general method be combined with automated sampletreatment in future studies. In addition, we argue that carefully estimated faecal volume production provides a simple and quick overall feeding estimate with important advantages over the common gut pigment technique, and it may be used as an independentmethod in bottle incubation experiments.