Plant-soil interactions shape how ecosystems respond to climate change, yet many of the underlying processes remain difficult to observe across scales and are challenging to represent in models. This session brings experimentalists and modellers together to examine how evidence from the plant-soil continuum can improve predictions of carbon, water and nutrient cycling from plots to whole ecosystems. We welcome studies from forests, grasslands, wetlands and permafrost systems that combine experiments, observations and modelling, or use one approach to challenge and guide the other. Contributions may address shared mechanisms across ecosystems, the translation of measured processes into model structure and parameters, and the design of experiments that can distinguish between competing model assumptions. By connecting these perspectives, the session aims to identify key knowledge gaps and practical ways to improve predictions of ecosystem responses to climate change.