Under conventional electron microscopy negatively stained phosphorylase kinase exhibits a bilobal structure resembling two bridged opposing parentheses. In this predominant particle orientation, usually only one bridge is observed; however, in many particles two bridges can be seen. Scanning transmission electron microscopy of unstained phosphorylase kinase shows very similar structures, with a particle mass equivalent to that of the hexadecameric holoenzyme. Partial digestion of the enzyme with chymotrypsin, which preferentially hydrolyzes the alpha-subunits, causes no significant changes in the structure; however, when both the alpha and beta subunits are degraded by trypsin, single lobed particles appear, i.e. the connecting bridges are missing. Mass analysis of scanning transmission electron microscopy images of trypsinized enzyme indicates that the protease does, in fact, split the particle into halves. Transmission electron microscopy of an alpha gamma delta complex isolated after incubation of the holoenzyme with LiBr shows only small particles approximately one-fourth the size of the holoenzyme. Thus, integrity of the beta subunit may be necessary in order for the two lobes of phosphorylase kinase to be bridged. These data also indicate that the subunits are arranged as a bridged dimer of octamers 2 (alpha 2 beta 2 gamma 2 delta 2).