A simple theoretical model of the process of strategic adaptation to climate change is proposed. Climate change is represented by a non-stationary Markov process on the space of climate states, and strategic adaptation by a simple resource allocation task in which agents incur costs when moving resources from one activity to another. A stationary analysis allows diagnostics that quantify the net costs of climate change, and the long-run benefits to adaptation, to be defined. A full dynamic analysis of the model allows for the computation of the costs of negotiating the transition between two stationary climate regimes. We analyze the dependence of these adaptation costs on the behavioural parameters of the model, and the costs of adjusting resources from one activity to another. We find that adaptation costs have a complex and counterintuitive dependence on adjustment costs, and can be more sensitive to the details of the climate change process than adaptation benefits are. This has important implications for adaptation planning, and understanding the linkages between adaptation and climate change mitigation.