The traditional approach to unraveling functions of a biochemical system is to study isolated enzymes and/or complexes, and to determine their kinetic mechanisms for catalyzing given biochemical reactions along with estimates of the associated param-eter values [51, 44]. While this reductionist approach has been fruitful, the buzzwords of the present are integration and systems. One of the important tasks in current com-putational biology is to assimilate and integrate the behavior of interacting systems of many enzymes and reactants. Understanding of such systems lays the foundation for modeling and simulation of whole-cell systems, a defining goal of the current era of biomedical science. In this paper we discuss approaches to modeling biochemical systems, with an emphasis on the basic concepts and techniques used in building large-scale integrated models of biochemical reaction networks. We consider the vices and virtues of the available methods; we speculate on what approaches are most reasonable for large-scale cellular modeling. How far current technology is from a reasonable quantification of whole-cell biochemistry depends on what level of detail one considers. At the simplest level (considering only reaction stoichiometry), whole-genome metabolic models of sev-eral single-celled organisms have been developed [2, 48, 23, 47, 52]. At the more detailed level of kinetic modeling, models of the relatively simple metabolism of the red blood cell represent some of the most ambitious attempts to date at modeling whole cell metabolism [24, 57, 28, 29].