Many published indexes in North American and Europe are based on mean or median prices, or applied repeat-sale method for less requirements of characteristic related information on properties. But each house is unique. It is critical that the differences among properties could be captured in the house price indexes in the timely and comparable way. This paper constructed daily apartment price indices for entire Sweden and 3 major cities - Stockholm, Gothenburg, and Malmö. We applied hedonic regressions on a unique cross-sectional time series database of over 1 million apartment transactions in Sweden from January 2005 to May 2015. Our newly constructed daily indexes compared time dummy variable approach, characteristics prices approach and the imputations approach. We tested spatial dependence, and performed spatial econometrics models for regional comparison among major cities.