Crystals of cytochrome bo(3) ubiquinol oxidase from E. coli diffract X-rays to 3.5 A and the structure determination is in progress. The limiting factor to the elucidation of the structural detail is the quality of the crystals; the diffraction spots from the crystals are diffused which leads to difficulties in processing the data beyond 4.0 A. Weak protein-protein contacts within the crystal lattice is assumed to be the cause of this problem. To improve these contacts, we have introduced protein Z to the C-terminal end of the subunit IV of cytochrome bo(3) and expressed both proteins as a single fusion. We have successfully obtained crystals of this fusion protein. The spot shape problem has clearly been solved in the crystals of the fusion protein although further optimization is necessary to obtain higher resolution. We also discuss the potential applications of this approach to the crystallization of membrane proteins in general.