1 Assoc. Prof. Ciprian Şipoş, Ph.D. West University of Timisoara, Faculty of Economic Sciences, Timişoara, Romania E-mail: ciprian.sipos@feaa.uvt.ro Eng. Adrian Crivii, FRICS, MBA Managing Partner, Senior Consultant Valuator, Darian Rom Suisse, Romania E-mail: acrivii@darian.ro JEL clasification: C20 A LINEAR REGRESSION MODEL FOR REAL ESTATE APPRAISAL Key words: real estate appraisal, statistical models, linear multiple regression Abstract: The real estate appraisal has an increasing importance due to strong growth of the real estate market in Romania in the last years. In that respect, a permanent preoccupation for specialists is to find newer and better methods to evaluate the real estates. Very used new approach of appraisal methods in the international practice is the statistical and econometric models. The main aim of the paper is to establish some principles of statistical real estate appraisal and to propose an applied linear regression model based on a panel data from city of Timisoara. The study areas covered by the statistical appraisal are selected from geographical localities, land categories or subjects to property taxes. The results of the appraisal model dramatically depend upon the extent and quality of available data. The information may be both quantitative – surfaces, types of buildings, improvements level, number of rooms and so on – or qualitative – the area accessibility, the market evolutions, the assessment of physical conditions. The model will be facilitated by computing methodology that includes multiple linear regression, statistical tests and variance analysis. The outputs of statistical appraisal are very important for the firms and government administration in the respect of obtaining an accurate view with minimal human and financial costs. The regression model provides a systematic application of appraisal methods to gain data for statistical review and analysis of the results. The model is capable to record adjustments in property categories, the level of values and consistency of assessments.