Multiscale simulations have been performed to address the longstanding issue of "dioxygen activation" by the binuclear copper monooxygenases (PHM and D beta M), which have been traditionally classified as "noncoupled" binuclear copper enzymes. Our QM/MM calculations rule out that Cu-M(II)-O-2(center dot) is an active species for H-abstraction from the substrate. In contrast, Cu-M(II)-O-2(center dot) would abstract an H atom from the cosubstrate ascorbate to form a Cu-M(II)-OOH intermediate in PHM and D beta M. Consistent with the recently reported structural features of D beta M, the umbrella sampling shows that the "open" conformation of the Cu-M(II)-OOH intermediate could readily transform into the "closed" conformation in PHM, in which we located a mixed-valent mu-hydroperoxodicopper(I,II) intermediate, (mu-OOH)Cu(I)Cu(II). The subsequent O-O cleavage and OH moiety migration to Cu-H generate the unexpected species (mu-O-center dot)(mu-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 O-2 activation by binuclear copper monooxygenases but also deciphers the full catalytic cycle of PHM and D beta M in accord with the available experimental data. These findings of O-2 activation and substrate hydroxylation by binuclear copper monooxygenases could expand our understanding of the reactivities of the synthetic monocopper complexes.