The purpose of this paper is to empirically verify the validity of the fundamentals model onresidential land prices from the view of long-run equilibrium and short-term dynamics. First, thenational panel data classified by municipal district were consolidated. Next, the equilibrium landprices were derived by panel cointegration analysis. Finally, the error correction models wereestimated, though some were not observed in the variables required for analysis. In surveying anddata consolidation, GIS (geographic information system) was utilized for the geographicaldistribution. Two versions of the aggregate land price data for every unit area were prepared theweighted average and the arithmetic average, and these were analyzed in comparison. In variouspanel estimations, the fixed effects model was adopted.A long-run equilibrium relation was observed by our fundamentals model as the result of panelcointegration analysis. Taxable income, expectation for future land prices, and real interest ratesgreatly contributed to the formation of the long-run equilibrium land price. The short-termfluctuation from the long-run equilibrium value appeared comparatively, notably in the municipaldistricts of the major urban areas. However, it was not necessarily a government building location inall prefectures.Next, the composition factor of the change rate of land prices was explored from the view of shorttermdynamics by carrying out a panel estimation of the ECM-type land price function. As a result ofexamining several models, we confirmed that, even if short-term deviation occurred in boththeoretical and actual land prices, about 60 percent of the deviation width is corrected the followingyear. These were in agreement with the quantitative consequence of the previous work.Many variables used for the verification of short-term dynamics were unstable for the parametersfor each model, or its significance. They also showed only a limited degree of incidence. On the otherhand, the population variable parameter showed significant and substantial influence in every model.In addition, the model by arithmetic average land price was supported by the weighted average landprice, with all the long-run equilibrium land price functions and the ECM-type land price functions,including additional analysis.