Chemotherapy of lung cancer is still an experimental approach requiring careful evaluation. Tumour response (marker of anticancer activity) is not perfectly correlated to survival (marker of chemotherapy efficacy), but its evaluation remains a milestone inasmuch as reporting a wrong tumour response rate might lead to the rejection of active new treatments. This review deals with the method of tumour response measurements and its use during a chemotherapy protocol. Recommendations drawn from the analysis of the literature are: 1) to assess and classify all lesions which can be identified at the beginning of the treatment; 2) to define the target lesions, mainly the ones which can be bidimensionally measured; 3) to use the World Health Organization recommendations for reporting the overall response; 4) to confirm complete response by negative rebiopsies; 5) to avoid second fiberoptic bronchoscopy to patients with stable or progressive disease on CT-scan, and finally; 6) to assess response quality by evaluating response duration and improvement of quality of life.