Agricultural and environmental systems are complex; being systems that can be configured in numerous ways, with many process based systems that interact in space and time, most under the influence of human intervention. For example, agricultural modelling started with simple representations (e.g. simulating growth and development of one crop in one dimension) and has taken on the challenge of representing increasing complexity. Some techniques for doing this aim to simplify the complexity to the dominant factors, other techniques attempt to expand the description of the system to encompass the complexity. Either path has implications for software to support the modelling. Papers within this session will encompass a wide range of contexts including parameterisation, model construction, software engineering, documentation, model testing, sensitivity or uncertainty or model-data assimilation.