Mathematical models are used to enhance our theoretical understanding of how population and community dynamics might behave under various ecological contexts, and statistical models are used every day to draw conclusions from empirical data. But the vast majority of our inference is based on the long-term, asymptotic properties of our models or static statistical relationships. Approaches for studying non-stationary ecological dynamics are needed now more than ever for predicting and forecasting how global change will affect the dynamics of wild organisms in decades to come. I will present small steps colleagues and I have taken to improve our ability to study the dynamics of wild populations in non-stationary environments created by global change.