Decision support systems (DSSs) rely on computational machinery in which mathematical models often constitute an important part. In this chapter, it is discussed which kinds of models are best suited for different kinds of DSSs. The practical steps involved in model construction are outlined, keeping in mind that model construction is a process that must be integrated into the larger software development project launched to construct the whole DSS. You are invited into the modeller’s workshop, as you follow the considerations involved in formulating a simple model of weed emergence. Two case studies close the chapter, demonstrating models of the population dynamics of annual weeds in a crop rotation and of an invasive weed. R scripts for all models can be found in the book’s online appendix. It is concluded that weed modellers must be prepared to work in multidisciplinary teams and that they should be better at considering the needs of the DSS users. For purposes of quality control, the mathematical models should be published open-source, while the DSS itself might be proprietary.