Blocking the terminal complement pathway with the C5 inhibitor eculizumab has revolutionized the clinical management of several complement-mediated diseases and has boosted the clinical development of new inhibitors. Data on the C3 inhibitor Compstatin and the C5 inhibitors eculizumab and Coversin reported here demonstrate that C3/C5 convertases function differently from prevailing concepts. Stoichiometric C3 inhibition failed to inhibit C5 activation and lytic activity during strong classical pathway activation, demonstrating a "C3 bypass" activation of C5. We show that, instead of C3b, surface-deposited C4b alone can also recruit and prime C5 for consecutive proteolytic activation. Surface-bound C3b and C4b possess similar affinities for C5. By demonstrating that the fluid phase convertase C3bBb is sufficient to cleave C5 as long as C5 is bound on C3b/C4b-decorated surfaces, we show that surface fixation is necessary only for the C3b/C4b opsonins that prime C5 but not for the catalytic convertase unit C3bBb. Of note, at very high C3b densities, we observed membrane attack complex formation in absence of C5-activating enzymes. This is explained by a conformational activation in which C5 adopts a C5b-like conformation when bound to densely C3b-opsonized surfaces. Stoichiometric C5 inhibitors failed to prevent conformational C5 activation, which explains the clinical phenomenon of residual C5 activity documented for different inhibitors of C5. The new insights into the mechanism of C3/C5 convertases provided here have important implications for the development and therapeutic use of complement inhibitors as well as the interpretation of former clinical and preclinical data.