When is it time to adopt different technologies, management strategies, and resource use practices as underlying climate change occurs? We apply risk and decision analysis to test hypotheses about the timing and pace of adaption in response to different profiles of climate change and extremes expressed as yield and income variation for a simulated dryland wheat farm in the United States Great Plains. Climate scenarios include gradual change with typical or increased noise (standard deviation), rapid and large change, and gradual change with extreme events stepped through the simulation. We test decision strategies that might logically be utilized by farmers facing a climate trend that worsens crop enterprise outcomes. Adaptation quickens with the rate of change, especially for decision strategies based on performance thresholds, but is delayed by larger climate variability, especially for decision strategies based on recognizing growing differential between adaptive and non-adaptive performance. Extreme events evoke adaptation sooner than gradual change alone, and in some scenarios extremes evoke premature, inefficient, adaptation.