Abstract. This paper deals with a prediction method using the hidden Markov model (HMM) for rain rate and rain propagation attenuation for K-band satellite communications links in tropical areas. As is well known, the K-band frequency is susceptible of being affected by atmospheric conditions, especially rain conditions. The wavelength of the K-band frequency, which approaches the size of a rain droplet, causes the signal strength to be attenuated and absorbed easily by the rain droplets. Therefore, in order to maintain the quality of system performance for K-band satellite communications links, special attention has to be paid to rain rate and rain propagation attenuation. Thus, a prediction method for rain rate and rain propagation attenuation based on HMM was developed to process measurement data. The measured and predicted data were then compared with ITU-R recommendations. The results indicate that the data show similarity with the model of recommendation ITU-R P.837-5 for rain rate and the model of recommendation ITU-R P.618-10 for rain propagation attenuation. Meanwhile, statistical data for measured and predicted data such as fade duration and interfade duration show insignificant discrepancies with the model of recommendation ITU-R P.1623-1.