Since the rapid urbanization period coincides with the start of global warming, climate data might be contaminated with urbanization effect in Korea. Monthly normals of daily minimum, maximum, and mean temperature of 14 stations were calculated for 1951-1980 and 1971-2000 periods. Differences in two temperature normals were regressed to the natural logarithms of population increase at corresponding 14 cities from 1966 to 1990. The regression models were used to remove potential effects of urbanization from the apparent warming, and to determine the net contribution of global warming to the temperature change in Korea during the recent half century. According to the model calculation, there was little evidence of global warming in the warm season (May through November), while urbanization effect was common in all season except April. Up to 0.5$^{circ}C$ warming of nighttime temperature was found to be induced by urbanization. Cool season temperature was increased by up to 0.6$^{circ}C$ due mainly to the global warming of daytime temperature.