In this paper, we present definitions of modularity and genericness that are based on sets of criteria and rules for model design and which encompass the goal of developing an efficient and flexible structure for plant and ecosystem models. Model structure should be based on modules that (1) relate directly to real world components or processes; (2) have input and output variables that are measurable values; and (3) communicate solely via these input and output variables. Such a model structure has the advantage that it can be incrementally improved by simply replacing one module with another that has the same input and output variables. The underlying mechanism in the replacement module can be different, which facilitates the incorporation of the latest experimental research results and allows modelers to readily test alternative hypotheses about mechanisms. Thus, modularity and genericness open models to contributions from many authors, facilitate the comparison of alternative hypotheses, and extend the life and utility of simulation models.