Professor Hendry investigates the theory and practice of econometric modelling and forecasting in a non-stationary and evolving world, when the model differs from the economy. First, model selection poses great difficulties, but recent research has revealed high success rates, and allows operational studies of alternative strategies. Second, when the processes being modelled are not time invariant, many of the famous theorems of economic forecasting no longer hold. A generalized taxonomy of forecast errors reveals the central role of unanticipated location shifts, and helps explain the outcomes of forecasting competitions. Surprisingly, other potential sources of forecast failure seem less relevant. Finally, co-breaking, corrections to reduce forecast-error biases, and model transformations all help robustify forecasts in the face of location shifts.