The real estate valuation in urban area is very difficult to make a good adjustment from many effecting qualitative factors. Market data approach relies on the market information and the reasonable adjustment with less subjectivity of the appraiser. This study tries to apply Fuzzy theory to deal with the subjectivity from the appraiser in determining the comparative of the qualitative variables. The qualitative factors considered in the real estate valuation model are first determined in the model using Quantification Theory I. The 22 effecting factors are considered in the model. And we make an evaluation formula in terms of the factors that show the maximum adjustment percentage to the real estate value from the market. The preliminary results are designed by a linguistic questionnaire and are surveyed by the officers in the fields of valuation. The fuzzy linguistic method is applied for solving the subjectivity of locational adjustment caused by the appraising officers. Integrated with the results and implemented with empirical evidence is established for the fair valuation of real estate for the final valuation model using Fuzzy Quantification Theory I. In order to verify the model we take 100 samples to review each case by the established model. The estimated results are demonstrated to show the differences. The final model is applied using the verified model. The qualitative variables are considered in the real estate valuation model and fuzzy linguistic model is applied to adjust the subjectivity of the appraisal. The modeling results are also verified by the samples that show the consistence of the appraising officers when using the fuzzy valuation model. The Tainan City case study shows some applicability of this model. The further study can be applied to a metropolitan area and the modeling parameters can be more realistic.