Existing viewpoints and theories of selective grazing by copepods are briefly reviewed in order to formulate explicit hypotheses to be tested experimentally. Based on these hypotheses, a series of grazing experiments was run to determine (1) the extent of the selective ingestion capabilities of Acartia clausi and (2) how these capabilities were affected by previous feeding histories. Groups of copepods were separately preconditioned on a small diatom (Thalassiosira pseudonana), a large diatom (T. fluviatilis), or a plastic sphere. The ingestive behavior was then examined on various combinations of spheres and food particles. Spheres offered alone were not ingested. In mixtures of diatoms and spheres, the copepods avoided ingesting spheres intermediate in size between the sizes of the diatoms. The copepods either ingested particles on either side of the spheres, or ignored all particles less than the size of the largest spheres. The pattern observed depended upon the size of the preconditioning food. However, if the spheres were larger than the largest food particles, the copepods still selectively ingested the food particles. The above results demonstrate that A. clausi has a complex grazing behavior consisting of (1) more efficient grazing on larger particles within its particle-size ingestion range; (2) the ability to alter “effective” setal spacing to optimize feeding behavior (i.e., the ability to increase efficiency of capture of food particles, and to avoid non-food particles); and (3) the ability for post-capture rejection of non-food particles when they interfere with the ingestion of food particles on which the copepod has been preconditioned. The behavioral patterns observed depend heavily on the food preconditioning and the presence or absence of non-food particles. These results clearly indicate that a simple “mechanistic” explanation of selective grazing is insufficient.