Hedonic regression models are commonly used tools in attempt to remove inherent heterogeneity in real estate market data. Our main objective is to develop hedonic time-series indices to be used in forecasting rent development in major Finnish regional markets. Another objective is to capture the development of appreciation for each explanatory variable in order to use that information to produce qualitative trend projections.Indices are constructed for 7 economic regions, which include 9 biggest cities in Finland. New leases data from KTI Finland is being used for modelling. The data comprises of c. 17,000 observations for period Sepí94-Augí06. Helsinki Metropolitan Area (HMA) is by far the biggest regional submarket with 11,000 observations. GIS software is being used for producing additional spatial variables. First approach is to create micro-indices for HMA and then theyíre aggregated into index for whole HMA region. Additionally cross-sectional location indices are created for HMA regional sub-markets. Cross-sectional coefficients are then used in foresight processes in order to extrapolate qualitative trends for real estate markets. Second approach is to create indices for each economic region both separately and by using panel estimation. Indices are tested for robustness and consistency.