To solve the problem of multiple continuous K nearest neighbor (KNN) queries over moving objects, considering the development of multi-core and multi-threading technologies, a two-stage framework is proposed for Multi-Threading Processing of Multiple Continuous KNN Queries (MPMCQ). This includes a preprocessing stage and a query execution stage to carry out the data updating task and the query execution task separately. In each of the stages, techniques are designed to optimize the cache access hit ratio and improve the parallelism through multi-threading. A query grouping technique in the query execution stage is proposed to improve the data temporal locality when accessing the memory. Thus, the cache hit ratio can be guaranteed. A KNN query algorithm is given based on the MPMCQ framework and the grid index for moving objects. Extensive experiments are carried out to verify that by adopting the multi-threading and the cache optimization technologies, the proposed framework implements a much superior performance than other famous algorithms; moreover, it maintains excellent performance scalability when executed under different multi-core CPUs.