Experiments on antibody feedback inhibition of the immune response have confirmed that control is more effective against a primary response than against a secondary response. The cells producing antibodies in primary and secondary responses are different both in terms of number of IgFC and amount of antibody produced by individual IgFC (plaque size). Late primary anti-burro RBC sera (greater than 200 days), despite low titers, are, on a volume for volume basis, feedback inhibitors at least as good as early (8-12 days) primary antisera on primary responses but are more effective in suppressing secondary responses (B memory cells). Late primary antisera, due to the process of affinity maturation, have a high affinity for antigen. The suppressive effect of early and late antisera is equally removable by absorption with burro erythrocytes: a result which it is thought, decreases the likelihood of feedback by anti-idiotype being involved in the observed suppression. It is suggested that feedback antibody acts (a) in competition with receptors, inter alia removing antigen into immunologically irrelevant pathways, (b) by a process involving the linking of antigen to Fc receptors, and (c) as a blocking antibody coating B cells (Bm) or APC which are already binding epitopes, thus preventing their cooperation with specific helper or other accessory cells.