Abstract: This study deals with grid-based outlier detection method which can figure out most outstanding outliers from a high speed datastreams. It is capable to find outliers even with the evolution of datastream where there is a chance that object properties may change with the time. Grid structure used in this study can help to save number of extra calculations in case of nearest neighbor queries and can provide a solid platform for applying distance based nearest neighbor approach for finding outliers. Proposed grid based method efficiently partition incoming stream into chunks and store these chunks one by one into a fixed width grid structure for further processing. Each chunk of stream is processed with the combination of fixed width grid structure and distance based nearest neighbor approach. Through efficient pruning of safe regions, proposed method only needs to operate over the candidate regions for finding outliers. This method takes into account both, local and global view of outliers and assign score to each detected outlier and does not sacrifice the correctness of its results for fast processing time. Proposed method can operate faster, need limited memory resources, having low computation cost and found to be highly efficient for data stream environment. Several experiments on real and synthetic datasets show the effectiveness of proposed method.