For the analysis of geothermal energy utilization in agriculture the relations between soil temperature and geographical variables such as latitude longitude and sea level in Korea were analyzed and the regression equations were suggested among them. The measured soil temperature data for four years in eighteen cities were used to get the soil temperature fitting equation depending on the soil depth and the time of year in each city. The mean correlation coefficient for those data fitting was 0.980. the correlation coefficient of regression analysis for the mean soil temperature($T_{m}$) on the geographical variables such as latitude longitude and height above sea level was 0.958 and those for soil surface temperature amplitude(Tss) and phase constant(tp) were 0.889, 0.835, respectively. The relation between the apparent thermal diffusivity of the soil and the three geographical variables was not significant. The regression equations for the mean soil temperature($T_{m}$) soil surface temperature amplitude(Tss) and phase constant(tp) adopting latitude($X_{1}$) longitude($X_2$) height above sea level($X_3$) were as follows : $T_{m}$=50.049 - $0.849X_1$-$0.03131X_2$-$0.00622X_3$Tss=-6.970 +$0.584X_1$+$0.00530X_2$-$0.00214X_3$tp=70.353 - $1.404X_1$+ $0.02098X_2$+ $0.00312X_3$