Unusual trends of surface temperature over recent decades have received much attention with an increasing interest in climatic changes due to natural causes and man-induced CO₂ effects. However, there have been suggestions that a proportion of the warming trend might be related to the urban effect. Therefore, to estimate the magnitude of urban bias quantitatively is considered vital for the detection and monitoring of possible long-term trends associated with increasing concentrations of atmospheric greenhouse gases. The purpose of this paper is to estimate the magnitude of urban bias in surface temperatures of South Korea using statistical procedures. Monthly mean temperatures of 16 stations were obtained for the period of 32 years(1968 ~ 1999). Each station is classified into urban and rural stations with its population densities, and urban stations are divided into large and smaller urban stations according to their population totals to examine magnitude changes of urban bias with the size of urban areas. Estimates of urban bias magnitude are calculated by averaging the difference between each urban station and every rural station. This difference is averaged over three groups of urban stations. Estimates of mean urban bias magnitude(T_u-r)are calculated by averaging the yearly urban bias estimates. Estimates of the urban trend(DT_u-r) are obtained using period means(doubling the differences obtained between yearly estimates averaged over two 16-year periods). For annual or seasonal mean temperature(T_i) the adjusted temperature (T'_i) is determined. As all estimates T_u-r of are greater than zero, it suggests that temperatures in urban stations are warmer than those in rural stations. Estimates of annual mean magnitude of urban bias range from 0.35°ĘC for smaller urban stations to 0.50°ĘC for large urban stations. These magnitudes are relatively smaller than those for other regions such as China and the United States. It might result from that rural stations sampled for this study have experienced more urbanization than rural stations used for those regions. Also, all estimates of DT_u-r are positive, indicating the increasing trend in urban bias time series. Seasonal variations are found and for urban stations, maximum T_u-r occurs in fall while minimum T_u-r in summer. It has been known that the intensity of urban heat islands is stronger fall and winter while weaker during summer in Korea. After adjusting the urban bias, an increasing trend in surface temperature series is still existed. However, it should be mentioned that results of this study can only be applied to correct possible urban bias not in individual station temperature series but in regional temperature series due to the large variation of urban bias magnitude. Also, none of rural stations used for this study can represent a true non-urbanized environment. Therefore, estimates of urban bias magnitude and trend from this study might be considered as measures of relative bias between heavily urbanized, industrialized areas and less urbanized areas.