Magnetotellurics has boomed in Australia over the last decade, from exploration-scale, industry-led surveys to lithospheric-scale, government-led initiatives like AusLAMP. However, key questions of how to interpret magnetotelluric models remain. Regardless of the scale of the survey, if magnetotelluric models are not interpreted in a way that will tell us more about the temperature and composition of the Earth, the value of those models to the wider geoscience community will always be restricted. In this talk, I will discuss the state of play of the tools we can use to interpret magnetotelluric models and I will show some recent developments in magnetotelluric interpretation. One key ingredient in those interpretations is the petrophysical and mineral physics data that link the electrical conductivity measured in magnetotellurics to rock composition and temperature. Another key is close integration with information from geology and other geophysical methods to ensure that interpretations are realistic and as tightly constrained as possible. I will discuss new experimental datasets and tools such as MATE that help MT practitioners access and implement these interpretations that are rooted in petrophysical and experimental mineral physics data. Focussing on recent interpretations of MT models from southern Africa and on compilations of MT signatures of kimberlites, I will demonstrate how advances in interpreting magnetotelluric models are having significant impacts on our understanding of lithospheric evolution and have the potential to guide mineral exploration strategies.