Accessing distant data items in mobile ad hoc networks (MANETs) environments pose a huge challenge due to mobility and resource constraints. A number of research studies have been developed to enhance data accessibility and decrease the number of pending queries. A collaborative caching technique is considered as an efficient mechanism to increase data accessibility and decrease the latency which in turn reduces the pending queries. However, the processing queries that based on their classifications along with distributed indexing have not been tackled in previous works to reduce the number of pending queries and increase the number of replied queries. A collaborative caching discovery which based on service providers (CCD) is proposed in this article, which process requests depending on their status either priority or normal. This ensures that the number of pending queries is reduced with minimum cache discovery overhead. The results of the experiment reveal that the proposed strategy increased collaborative caching discovery efficiency and outperformed the cooperative and adaptive system (COACS) in terms of increasing the number of replied queries and reduction of the pending queries by 24.21 percent.