This article explores the practice of simulation modeling byinvestigating how parameterizations are constructed and integratedinto existing frameworks. Parameterizations are simplified processdescriptions adapted for simulation models. On the basis ofa study of meteorological research, the article presents predictiveand representative construction as two different ways of developingparameterizations and the trade-offs involved in this work.Because the overall aim in predictive construction is to improveweather forecasts, the most practical solutions are chosen overthe best theoretical solutions. In representative construction,the situation is reversed, but while discourse focuses on theoryand models, the everyday work is often tied to computer programs.These different ways of construction work are closely relatedto the role of the simulation models as epistemic or technicalobjects, and this characterization is also used to compare theresults with previous research.