Notwithstanding the additional spectrum allocations of the 1979 World Administrative Radio Conference (WARC '79) and the ever-increasing sophistication of spacecraft antennas which allow more intensive spectrum reuse, saturation of the 6/4 and 14/11 GHz bands is foreseeable by the late-1990s.The next available frequency bands at 30/20 GHz were allocated at WARC '71, when knowledge of propagation phenomena and their effect on satellite communications systems performance was fragmentary. During the past 8 years, numerous experimental programmes have led to the acquisition of more precise information about hydrometeor-induced propagation disturbances. Analysis of the data reveals that even with corrective means such as diversity and coding, service reliability of satellite communications systems with up-links at 30 GHz and down-links at 20 GHz could not reach 99.99 per cent as currently available at the lower frequencies and could be 98.0 per cent, or lower, in certain parts of the world. Since the conventionally proposed use of the 30/20 GHz bands does not lead to optimal solutions, alternative approaches are needed.An approach which would involve a revision of the spectrum allocations at 30/20 GHz is proposed. Placing both up-and down-links around 20 GHz and redistributing the spectrum available at 20 GHz would prevent the intrinsic imbalance of the up-and down-links with consequent higher reliability, reduced earth-station transmitter power, reduced construction and operational costs, and simplification of the spacecraft transponders and antenna design.The use of frequencies around 30GHz could be postponed until the 20GHz bands became saturated. When the 30 GHz bands would eventually be used, there would be no imbalance of up-and down-links .