Low frequency (decadal and longer) climate variability and anthropogenic climate change pose a significant challenge to the management and planning of water resources infrastructure. Both climate aspects lead to some common institutional challenges for design, operation and policy formulation or implementation. How should water allocation and reservoir operation rules be changed if we expect a systematic change in the long term mean, variance and other statistics of streamflow, but are faced with considerable uncertainty as to the nature and magnitude of this change? What are the implications of constructing additional reservoir storage or the removal of a dam given that the next 20 or 30 years may see a significant change in flow attributes? How does one interpret climate variability in the past several centuries in the context of what can be allocated and when, and to assess the frequency, severity and duration of supply or allocation failures? What aspects of these multi-decadal system performance attributes are modified by reservoir storage of a particular size, and which risks remain residual to storage provision? These types of questions are of great interest for the Colorado River Compact. As is well known, the Compact allocation was based on an anomalously wet period, which was followed by multiple decades with well below average flow. Multi-century Colorado River streamflow reconstructions show preferred regimes with recurrence structure or cycles, whose role needs to be understood in the context of managing climate risk for the long run performance and adaptive application of the existing Colorado River Compact, or for the modification of the Colorado River Storage Project's components or operating rules. Here, we use more than 5 centuries of reconstructed streamflow to assess some aspects of the dynamic climate risk faced by this system, with a specific focus on the example questions listed above. We derive the performance attributes of the system through simulations of the Glen Canyon reservoir's mass balance, and explore the time-frequency structure of reservoir storage and supply/deficit in supply from the reservoir. We note that the Glen Canyon reservoir is effective in ameliorating the impacts of inter-annual and decadal climate variability, but multi-decadal climate risk, as often discussed in the context of the history of the Colorado River Compact is not significantly reduced. The prospects for developing a strategy to identify the operative climate regime, and the potential scenarios for the short, medium and long run inflows, and the use of these scenarios to develop an adaptive management strategy for the system are discussed.