The present invention relates to methods for predicting the survival time of patients suffering from a lung cancer. In particular, the present invention relates to a method for predicting the survival time of a subject suffering from a lung cancer comprising the steps of i) quantifying the density of regulatory T (Treg) cells in a tumor tissue sample obtained from the subject, ii) quantifying the density of one further population of immune cells selected from the group consisting of TLS-mature DC or TLS-B cells or Tconv cells, CD8+ T cells or CD8+ Granzyme-B+ T cells in said tumor tissue sample, iii) comparing the densities quantified at steps i) and ii) with their corresponding predetermined reference values and iv) concluding that the subject will have a short survival time when the density of Treg cells is higher than its corresponding predetermined reference value and the density of the further population of immune cells is lower than its corresponding predetermined reference value or concluding that the subject will have a long survival time when the density of Treg cells is lower than its corresponding predetermined reference value and the density of the further population of immune cells is higher than its corresponding predetermined reference value.