The basic characteristics of radio paths relevant to mobile radio communication at UHF are reviewed. Data gathered in a comprehensive survey of urban and suburban Melbourne were analysed to test for consistency with several simple models: Rayleigh envelope, a power spectrum model, log-normal distribution of mean, and the effects of distance and slope of terrain. Factors limiting the validity of some of the models are discussed. Other factors investigated were repeatability under varying weather conditions, the averaging required for reliable representation of local field strengths, and distributions associated with slow and fast fading patterns. A study of the applicability of space diversity showed that a saving of at least 75% of the transmitter power, for a given grade of service, could be made if diversity were used. Because of the number and complexity of the factors influencing the propagation to a vehicle, and because weather and time of day were not found to be significant influences, it was concluded that experimental assessment is a cost-effective method for determining accurately a service area. Examples of suitable data gathering phase are given.