caused by mist events and mist and rain events (which could not be separated). The highest attenuation part of the distribution follows a -1/2(dB/km)/decade slope for the low percentages of time.Highest attenuation values recorded during the measurement period were never > 1 3 d B h , which is comparable with the results of Chu and Hogg [l] for a 2.6km, 0 . 6 3~ path. Fig. 2 presents the cumulative distribution of attenuation for rainfall/ mist events only. The highest attenuation part of the distribution follows a 2/3 (dB/km)/decade. The season for greatest attenuation was found to be the winter, when mist, very often accompanies rainfall. tiles of the distributions were obtained by fitting curves to the empirical distributions. A piecewise linear fit was used for the distribution of attenuation, and an exponential fit was used for the distribution of rainfall. Fig. 4 shows the diagram of the percentiles of attenuation against the percentiles of rainfall rate obtained in this way. The values of the parameters a and b obtained were quite different to those obtained by Gibbins et al. [3], which were -2 and 0.6, respectively. However, their measurements were for rainfall only. The discrepancy is to be expected as the results are not comparable. The attenuation measured in this work is not only caused by rainfall but includes the very practical case of the presence of mist. Also, the single measurement of rainfall rate is not necessarily indicative of the rainfall rate along the path.Conclusions: Cumulative distributions of attenuation and rainfall rate were obtained for one year of measurements on a 4.1 km line of sight link operating at 1 . 5 5~. Winter was found to be the worst season for attenuation, in part caused by the mist occurring simultaneously to rain.