Traditional plant architectural models or 'visualization models' propose to visually create realistic three-dimensional plants. The visualization is based on field sampling and the application of an algorithm to standardize the three- dimensional description of a plant. "L-systems" and the "Reference Axis" are two such approaches. Mechanistic or physiologically based models, built using mathematical expressions of the interactions between plant components, propose to describe how a plant functions. They simulate physiologically realistic plants based on estimates of physiological development and growth. Their equations are derived from field experiments. In this study we integrated both modeling paradigms. We used functions and concepts obtained from mechanistic and architectural modeling theories and developed an integrated system. The system was derived from an enhanced 'mechanistic' model, GOSSYM, with 3D architectural extensions. We accomplished this by associating growth and development functions with actual locations in three-dimensional space. The resulting model allows vastly improved model output interpretation, use of the model as a surrogate experimental environment and allows better integration of our knowledge about how plants grow into a unique system. The new model, named COTONS, produces "life-like" plants. Now the farmer deals with simulation results analogous to the ones he deals with in a 3 -dimensional world. Very importantly, variability is captured and expressed visually. This is the first step for better characterizing production risk in humanbased terms. This new model symbolizes crop models for the next century.