The reductionistic approach to cancer research has led to an enormous amount of information and publications regarding the molecular biologic processes that take place in cancer tissue. However, the specific influence of this information on clinical practice has been limited. With the advent of new reductionistic tools like the transcriptomic and proteomic technologies, many would argue that further advances in the field of lung cancer research will be dominated by advances on the technical level. However, we anticipate that the most revolutionary advances will be those at a conceptual level. Medical science has always been reductionistic in essence, reducing and analyzing the composing elements of our complex biologic machinery, overlooking the fact that the interrelation among a set of simple determinants creates a new dimension of characteristics and functions. Problems emerging from a reductionistic approach are heterogeneity and variability. This review addresses the current conceptual problems in the field of lung cancer biology and provides a new conceptual model based on recent publications.