This study examines various methods for correcting the smoothness of appraisal-based indices proposed by real estate research. Such methods are widely applied in asset allocation decisions when transaction-based indices are not available. Using more than 40 years of index data, we investigate the distributional features of the obtained “unsmoothed” return series on both a quarterly and an annual basis. Repeated evaluations with evolving time windows reveal a remarkable lack of distributional stability for the majority of techniques. We also propose and test modified procedures for several models. Our results suggest a need for caution when applying appraisal-based real estate indices in an asset allocation context, as well as for interpreting the empirical results from unsmoothed data. The results suggest that less sophisticated correction techniques may provide equivalent solutions to address the smoothness of appraisal-based returns.