•We analyzed the outcomes of 342 lung cancer patients receiving salvage chemotherapy after immunotherapy (SCAI).•SCAI obtained a median overall survival (OS) of 6.8 months (95 % CI 5.5–8.1).•The median progression-free survival (PFS) was of 4.1 months (95 % CI 3.4−4.8).•The objective response rate (ORR) was of 22.8 %.•A “Post-CKI score” was constructed by combining significant predictors of survival. In the most of cases, for non-small cell lung cancer (NSCLC) patients who progressed to previous immune checkpoint inhibitors (CKI) administered as first- or as second-line therapy, chemotherapy (CT) remains the only viable options in the absence of “druggable” mutations. We aimed to explore the efficacy of salvage chemotherapy after immunotherapy (SCAI) in advanced NSCLC patients. We designed a retrospective, multicenter study, involving 20 Italian centers, with the primary objective of describing the clinical outcome of advanced NSCLC patients treated with SCAI at the participating institutions from November 2013 to July 2019. The primary endpoint of the study was represented by overall survival (OS), defined as the time from CT initiation to death. Secondary outcome endpoints of the SCAI (progression free survival, PFS, objective response rate, ORR and toxicity) and explorative biomarkers (lactate dehydrogenase, LDH, and neutrophil-to-lymphocyte ratio, NLR during immunotherapy) were also analyzed. In our study population of 342 NSCLC patients, SCAI obtained a median OS of 6.8 months (95 % confidence interval, CI 5.5–8.1), median PFS of 4.1 months (95 % CI 3.4−4.8) and ORR of 22.8 %. A “Post-CKI score” was constructed by combining significant predictors of OS at the multivariate analyses (sex, ECOG PS, disease control with prior immunotherapy), Harrell’C was 0.65, (95 % CI:0.59−0.71). Despite the late-line settings, our findings support the hypothesis that previous immunotherapy might increase the sensitivity of the tumor to the subsequent chemotherapy. The “Post-CKI score” was clinically effective in successfully discriminating three distinct prognostic subgroups of patients after the failure of CKI, representing a possibly useful tool for the tailored decision-making process of advanced treatment-line settings in NSCLC.