We present a replica placement scheme for any distributed hash table that uses a prefix-matching routing scheme and evaluate the number of replicas necessary to produce a desired number of disjoint routes. We show through simulation that this placement can make a significant improvement in routing robustness over other placements. Furthermore, we consider another route diversity mechanism that we call neighbor set routing and show that, when used with our replica placement, it can successfully route messages to a correct replica even with a quarter of the nodes in the system failed at random. Finally, we demonstrate a family of replica query strategies that can trade off response time and system load. We present a hybrid query strategy that keeps response time low without producing too high a load.