Most trials with Immune Checkpoint Inhibitors (ICIs) for Non-Small Cell Lung Cancer (NSCLC) included only small subgroups of patients (pts) aged ≥65. As NSCLC is often diagnosed in pts aged ≥70, real-world data about efficacy and safety of IO in elderly pts are essential. We retrospectively collected data about all pts with advanced NSCLC treated with IO at our Institution between April 2013 and March 2019. All ICIs administered for ≥1 cycle were admitted. Pts were stratified for age as follows: <70 year-old (yo), 70-79 yo, ≥80 yo. Chi-square test was used to compare qualitative variables. Survival was estimated with Kaplan-Meier method. Log-rank test was used to compare curves. Multivariate analyses were performed with Cox model. We reviewed 290 cases, with a median age of 67 (range: 29-89). Pts aged<70, 70-79 and ≥80 yo were 180, 94 and 16, respectively. Two hundred five pts received an anti-PD1, 77 an anti-PDL1, 8 an anti-CTLA4 or a combo-IO. Clinical/pathological variables were uniformly distributed across age classes, except for a higher rate of males (p 0.0228) and squamous histology (p 0.0071) in the intermediate class. Response Rate (RR) was similar across age groups (21.5% vs 22.3% vs 18.8% for pts aged<70 vs 70-79 vs ≥80 yo, respectively; p 0.9470). Median PFS did not differ according to age (2.8 vs 3.5 vs 2.6 mos for pts aged<70 vs 70-79 vs ≥80 yo, respectively; p 0.2020). Similarly, median OS was similar across age classes (9.1 vs 11.3 vs 9.6 mos for pts aged<70 vs 70-79 vs ≥80 yo, respectively; p 0.9144). These results did not change after stratification for sex (p 0.516 for PFS, p 0.5154 for OS) and histology (p 0.9057 for PFS, p 0.1002 for OS). The incidence of toxicity was comparable across subgroups (grade ≥2 adverse events in 35.8% vs 32.7% vs 37.5% for pts aged<70 vs 70-79 vs ≥80 yo, p 0.6493). The only variables influencing outcome at both univariate and multivariate analyses were performance status (p<0.0001 for PFS, p 0.0192 for OS), number of metastatic sites (p 0.0842 for PFS, p 0.0235 for OS) and IO line (p<0.0001 for both PFS and OS), regardless age group. Advanced age is apparently not associated to a reduced efficacy of IO in our case series. Furthermore, no toxicity concern emerges even among the eldest pts. Therefore, to our opinion ICIs should be considered irrespective of age, provided an optimal PS at baseline. Of note, IO is often the only therapeutic option applicable to these cases considering the toxicity of chemotherapy.