Multiscale simulations have been performed to address the longstanding issue of "dioxygen activation" by the binuclear copper monooxygenases (PHM and DβM), which have been traditionally classified as "noncoupled" binuclear copper enzymes. Our QM/MM calculations rule out that CuM(II)-O2• is an active species for H-abstraction from the substrate. In contrast, CuM(II)-O2• would abstract an H atom from the cosubstrate ascorbate to form a CuM(II)-OOH intermediate in PHM and DβM. Consistent with the recently reported structural features of DβM, the umbrella sampling shows that the "open" conformation of the CuM(II)-OOH intermediate could readily transform into the "closed" conformation in PHM, in which we located a mixed-valent μ-hydroperoxodicopper(I,II) intermediate, (μ-OOH)Cu(I)Cu(II). The subsequent O-O cleavage and OH moiety migration to CuH generate the unexpected species (μ-O•)(μ-OH)Cu(II)Cu(II), which is revealed to be the reactive intermediate responsible for substrate hydroxylation. We also demonstrate that the flexible Met ligand is favorable for O-O cleavage reactions, while the replacement of Met with the strongly bound His ligand would inhibit the O-O cleavage reactivity. As such, the study not only demonstrates a "coupled" mechanism for O2 activation by binuclear copper monooxygenases but also deciphers the full catalytic cycle of PHM and DβM in accord with the available experimental data. These findings of O2 activation and substrate hydroxylation by binuclear copper monooxygenases could expand our understanding of the reactivities of the synthetic monocopper complexes.