Immune checkpoint inhibitors (ICI) have revolutionized the management of advanced non-small cell lung cancer (NSCLC). However, most pivotal phase III trials systematically excluded patients with active brain metastases, precluding the generalization of the results. Although theoretically restricted from crossing the blood-brain barrier, the novel pharmacokinetic/pharmacodynamic profiles of anti-PD-1/PD-L1 drugs have prompted studies to evaluate their activity in patients with NSCLC with active central nervous system (CNS) involvement. Encouraging results have suggested that ICI could be active in the CNS in selected patients with driver-negative advanced NSCLC with high PD-L1 expression and low CNS disease burden. Single-agent CNS response rates around 30% have been reported. Beyond this particular setting, anti-PD-1/PD-L1 antibodies have been evaluated in patients receiving local therapy for brain metastases (BM), addressing concerns about potential neurologic toxicity risks associated with radiotherapy, more specifically, radionecrosis (RN). Accordingly, a variety of clinical and imaging strategies are being appropriately developed to evaluate tumor response and to rule out pseudoprogression or radionecrosis. Our purpose is to critically summarize the advances regarding the role of systemic anti-PD-1/PD-L1 antibodies for the treatment of NSCLC BM. Data were collected from the PubMed database, reference lists, and abstracts from the latest scientific meetings. Recent reports suggest anti-PD-1/PD-L1 agents are active in a subset of patients with NSCLC with BM showing acceptable toxicity. These advances are expected to change soon the management of these patients but additional research is required to address concerns regarding radionecrosis and the appropriate sequencing of local and systemic therapy combinations.