Similarity search in time series databases is an important research direction. Several methods have been proposed in order to provide algorithms for efficient query processing in the case of static time series of fixed length. In streaming time series the similarity problem is more complex, since the dynamic nature of streaming data make these methods inappropriate. In this paper, we propose a new method to evaluate similarity range queries in streaming time series. The method is based on the use of a multidimensional access method, the R tree, which is used to store features of the time series extracted by means of the DFT (Discrete Fourier Transform). We take advantage of the incremental computation of the DFT and equip the R -tree with a deferred update policy in order to improve maintenance costs. The experimental evaluation based on synthetic random walk time series and on real stock market data shows that significant performance improvement is achieved in comparison to the sequen...