Relative humidity (RH) is one of the climate elements representing the amount of moisture in the atmosphere, so the RH is defined as the ratio (%) of the actual water vapor pressure to the saturated vapor pressure for the given temperature. Study on the RH is important for understanding changes in the hydrologic cycle including the formation of cloud and fog. In this study, the spatio ? temporal characteristics, long-term trends, and regional types were analyzed based on statistical characteristics of observed RH in Korea. To analyze the detailed spatio ? temporal distribution and variation characteristics, 454 observation data from the KMA for the recent 5 years (2015.12 ? 2020.12) were used. The quality of raw RH data was improved through the quality control process. Inverse distance weighting method was used for the generation of 1km gridded RH data. RH shows strong seasonal variations, high (> 70%) in summer and autumn, and low (< 65%) in winter and spring, regardless of geographic location. Compared to other seasons, there are strong spatial variations in winter and spring, with very low RH but high coefficient of variation (CV) in the eastern part of the Taebaek Mountains. The RH for the recent 30 years (1990.12~2020.12) showed a de-creasing trend in winter and spring, but an increasing trend in summer and au-tumn. The decreasing trend of RH in winter is mainly due to a decrease in the dew point temperature, whereas in the case of spring, it is due to a rapid in-crease in temperature. In addition, the increasing trend of RH in summer and au-tumn is closely related to the increase in the dew point temperature. The K-means method was used for clustering based on time-varying characteristics (seasonal average, CV, ..) of RH. The 67 RH data were classified into 6 clusters, and there were significant differences in the seasonal average, CV and trend among clusters. Cluster 2 (mainly located at western boundary) has the highest RH as a whole and the smallest CV, showing a particularly increasing trend. On the other hand, group 6 (Seoul and Daegu) has the lowest RH and a large CV and decreasing tendency. Also, there is a large difference in the seasonal variation pattern according to the cluster, such as the magnitude of seasonal variation in RH and the time when the maximum and minimum RHs occur.