Different structures are used in peer-to-peer networks to represent their inherently distributed, self-organized and decentralized memory structure. In this paper, a simple rangequeriable distributed data structure, called RAQ, is proposed to efficiently support exact match and range queries over multi-dimensional data. In RAQ, the key space is partitioned among the network with n nodes, in which each element has links to O(log n) other elements. We will show that the look-up query for a specified key can be done via O(log n) message passing. In addition, RAQ handles range-queries in at most O(log n) communication steps. We also present a probabilistic variation of RAQ in which each node has O(log log n) links on the average and look up and range queries are resolved with O(log2 n) latency. At the end we evaluate RAQ through simulation.