Monoclonal antibodies were used to determine the number and molecular form of C3 bound to particulate activators of the complement (C) system by human serum. Sheep erythrocytes (E) coated with IgM (EIgM) and IgG (EIgG) were used to study activation of the classical pathway (CP). Yeast (Y), rabbit erythrocytes (ER), and five species of bacteria (Escherichia coli, Staphylococcus aureus, Streptococcus pneumoniae type 3, Streptococcus pyogenes, and Hemophilus influenzae type b) were used to study activation of the alternative pathway (AP). The deposition of C3b onto EIgM and EIgG incubated in C7-deficient human serum was dependent on the serum concentration. At all serum concentrations tested, there was complete conversion of C3b to iC3b. Kinetic analysis of C3b deposition and conversion to iC3b indicated that these events occurred almost simultaneously; the reaction was completed by 15 min. The deposition of C3 onto the AP activators ER and Y was also dependent on serum concentration, and ER, but not Y, required the presence of Mg-EGTA and thus the activation of only the AP. C3b deposition and conversion to iC3b on Y was complete in 15 min, with 82% of bound C3 converted to iC3b. For ER, maximum C3 deposition required 30 min in both the presence and absence of Mg-EGTA. However, after 1 h of incubation, 74% of bound C2 was iC3b in the absence of Mg-EGTA, compared with only 52% in the presence of Mg-EGTA. Thus, even on AP activators, a large portion of C3b may be converted to iC3b, and this conversion is probably controlled by elements on the particle's surface. Studies with the five species of bacteria yielded similar results. Approximately 3-5 X 10(4) molecules of C3 were bound per microorganism, with opsonization being completed in 30 min. Remarkably, only 16-28% of bound C3 was in the form of iC3b, even after 2 h of incubation. The presence or absence of Mg-EGTA, or the addition of purified CR1 to the reaction mixture, did not significantly effect the ratio of C3b to iC3b. Finally, SDS-PAGE and autoradiography of particle-bound 125I-C3 fragments confirmed that there was no conversion of iC3b to C3d,g or C3d. The data obtained about the opsonization of bacteria suggest that the predominant form of C3 that is encountered by inflammatory phagocytes may be C3b.