The outlook for patients with lung cancer remains poor despite advances in the understanding of the pathology and biology of this disease. To optimize treatment protocols prognostic data are essential. The current era with molecular research on mRNA expression analysis and proteomics will lead to a plethora of new molecular markers, which are likely to be correlated, at least in part, with each other and with disease activity, progression and survival. However, although the number of prognostic factors analysed in published systematic reviews on lung cancer is large, the scope of these factors in individual studies is often narrow. In daily practice prognostic factors other than general TNM staging are not implemented. To assess the efficacy of new prognostic factors for the management of individual patients with non-small cell lung cancer, studies with clinically relevant modelling are required. In this review arguments are provided to use a model combining radiological and histopathological growth rate, histopathological diagnosis and molecular characteristics as markers for metastatic capacity, tumour volume doubling time and expected response to targeted therapy. This may reveal time-related predictive information useful for treatment guidance of the individual patient.