Introducing a Mobility Anchor Point (MAP), Hierarchical Mobile IPv6 (HMIPv6) reduces the signaling overhead and handoff latency associated with Mobile IPv6. However, if a Mobile Node (MN)â„¢s session activity is high and its mobility is relatively low, HMIPv6 may degrade end-to-end data throughput due to the additional packet tunneling at the MAP. In this paper, we propose an Adaptive Route Optimization (ARO) scheme to improve the throughput performance in HMIPv6 networks. Depending on the measured Session-to-Mobility Ratio (SMR), ARO chooses one of the two different route optimization algorithms adaptively. Specifically, an MN informs a Correspondent Node (CN) of its on-Link Care-of Address (LCoA) if the CNâ„¢s SMR is greater than a predefined threshold. If the SMR is equal to or lower than the threshold, the CN is informed with the MNâ„¢s Regional CoA (RCoA). We analyze the performance of ARO in terms of balancing the signaling overhead reduction and the data throughput improvement. We also derive the optimal SMR threshold explicitly to achieve such a balance. Analytical and simulation results demonstrate that ARO is a viable scheme for deployment in HMIPv6 networks.