Geostatistics and spatial econometrics are two spatial statistical approaches used to deal with spatial dependence. Geostatistics estimates directly the variance-covariance matrix by assuming that the covariance among observations depends inversely on the distance between their locations, called the covariogram. Spatial econometrics defines and integrates the spatial interaction matrix in a hedonic regression model. In real estate, price estimation should take into account these spatial characteristics because property prices are correlated. Hence, these two approaches are commonly used to study the spatial dependence of the real estate prices in many contexts. However, a definite rule in selection these statistic approaches has not been established. This thesis examined these two approaches in order to distinguish the similarities, differences, advantages, and disadvantages of each methodology. Some examples of their applications in a real estate study. The geostatistics is used to analyze the stationarity of the variogram and its sensitivity depending on the parameters added in hedonic estimation. The spatial econometric is used to define econometrically the real estate market dominant area
