By exploiting query predicate similarities in the data streams, queries optimization is a mainly strategy to share computation. In this paper, based on the sliding windows, an approach is presented for the multiple range aggregation queries over spatio-temporal streams. Time fragments were employed to describe the overlaps between sliding windows and spatial fragments for the overlaps between spatial range predications. Then a technique was proposed to partition the streams into different spatio-temporal fragments without any overlaps. The final query results could be captured from the temporary query results over the spatio-temporal fragments. Experimental result shows that our approach can substantially increase performance of the multiple range aggregation queries over spatio-temporal streams.