Résumé : Intensity modulated radiotherapy is increasingly used in non-small-cell lung cancers despite a low level of evidence. A literature review was conducted. Several critical physical and dosimetric uncertainties are however unsolved. Methods to circumvent these limitations are being developed. In several retrospective studies, survival rates were at least similar with intensity-modulated radiotherapy as those reported with three-dimensional irradiation. To date, intensity modulated radiotherapy might be authorized in complex anatomical situations such as tumours close to the spinal cord (such as Pancoast Tobias, paraspinal and paracardiac tumours) or with limited motion amplitudes. Dosimetric benefits should also account for 4D dose distribution issues. The reduction of intermediate and high doses in the organs at risk with intensity modulated radiotherapy is advantageous. However, the effect of low doses in large volumes (lung, bone, unspecified tissues along beam paths) and the effect of increasing integral dose are still poorly known. In conclusion, dose-volume correlations need to be better documented and prospective randomized trials should be encouraged.