A spread-spectrum technique for cellular high-capacity mobile communications is described and some results from an analytic study are summarized. The technique uses a very large set of frequency-hopped signals which are designed for minimal mutual interference. No synchronization of the mobile units is required, and each user is permanently assigned his own signal, which serves as an identifying signal and as a carrier for the biphase-modulated digital message. The spectral efficiency of the spread-spectrum system is analyzed and compared with the efficiencies of developmental FM/channel reuse schemes currently under construction in the U.S. and Japan. It is concluded that even with relatively simple speech digitization schemes, the efficiency of the spread-spectrum scheme may exceed those of the narrow-band schemes by a factor of almost five. More ambitious bit-rate-reducing speech digitization methods could improve still further on these figures. Additional benefits of the spread-spectrum scheme include immunity from fading and interference, more consistent speech quality, simpler system control algorithms, and more flexible blocking properties under overload conditions.