Developing tools at farm and regional levels faces a double methodological challenge: how to up-scale biological knowledge from the animal/plant level to higher levels of organization and how to avoid the pitfall of labyrinthine models, not easy to use and disseminate? In other words, how can we find the appropriate level of simplification when developing models and tools to understand and evaluate complex systems? Part of the answer implies to identify the key drivers and effects that should be conserved when changing the level of organization. It is also important to clarify input/output ports that connect levels and therefore ensure that information is transmitted across scales with relevant accuracy. Considering these aspects is also crucial when communicating the complexity of livestock systems and transferring knowledge to end-users. Based on concrete examples of models and tools, this session illustrates how it is possible to get the right level of simplification when working in more and more aggregated levels of organization and how to communicate complex relationships. A short introduction with two examples of models combining genetics, nutrition and economics will start the session, followed by four communications. Based on these various examples, the session will end with a general discussion of challenges and future research needs on tools/models for LFS research.