INTRODUCTION: The present study aimed at comparing the efficacy in recognizing depression, in 53 patients newly diagnosed with lung cancer, of the Hospital Anxiety and Depression Scale (HADS), a self-report screening questionnaire, and the Montgomery-Asberg Depression Rating Scale (MADRS), a semi-structured clinician-rated interview. Specifically, we aimed at addressing the question of which is the best HADS cutoff for the detection of patients to be further investigated through a clinical semi-structured interview (the MADRS). RESULTS: The MADRS identified 92% of the patients as depressed; the HADS, 70% and 87%, using a cutoff of 11 and 8, respectively. The results suggest substantial agreement between the HADS and the MADRS when a cutoff of 8 is used (McNemar: p = 0.51; Cohen K = 0.69), while a HADS cutoff of 11 resulted in a significantly lower concordance with the MADRS (McNemar: p = 0.002; Cohen K = 0.49).