C-ros proto-oncogene-1 (ROS1) chromosomal translocations result in constitutively activated receptor tyrosine kinase in 1-2% of NSCLC. Although crizotinib, a first-generation Tyrosine Kinase Inhibitor (TKI) of Anaplastic Lymphoma Kinase (ALK) and ROS1 translocations, has demonstrated efficacy in treatment of ROS1 NSCLC, patients invariably acquire resistance mutations, with a high incidence of brain metastasis at relapse owing to poor CNS drug penetration. Lorlatinib is a novel, third-generation TKI with improved CNS penetration and preclinical activity against crizotinib-resistance mutations, which recently demonstrated efficacy in a phase I/II trial. We report a patient who developed secondary crizotinib-resistance and was successfully treated with lorlatinib. A 73-year-old, female, never-smoker with recently diagnosed stage-IV ROS1-rearranged NSCLC presented with refractory seizures and coma. The patient was on first-line treatment with crizotinib for 6-months, with an excellent systemic response at 2 months. Brain MRI after presentation demonstrated multiple intracranial metastasis with vasogenic edema (Fig.1A). Neurologic status was unchanged despite mannitol, steroids and anti-epileptics. As salvage therapy, the patient received bevacizumab for vasogenic edema and lorlatinib (100mg/daily). After 48-hours, substantial clinical improvement in neurologic function and mental status was observed. A repeat MRI (D+4) demonstrated a slight interval decrease in the largest metastasis and associated edema. At 1-month follow-up, the patient continued to improve clinically with no evidence of lorlatinib toxicity. Brain MRI showed a decrease in the sizes and/or resolution of all previously noted lesions and no new metastases or significant peritumoral edema (Fig.1B). Since clinical trials typically exclude patients with symptomatic, untreated CNS disease, efficacy of newer, targeted agents in mutated-NSCLC is unclear. Here we report on a patient with life threatening, acute neurologic deterioration who derived a meaningful clinical reduction in CNS lesions at 1 month, with favorable CNS response at 2 months. In addition, we observed early evidence of tumor regression on MRI after 4 days of lorlatinib therapy. Of note, her rapid clinical response may have in part been attributable to the addition of bevacizumab. Another important consideration for our patient is that she avoided whole-brain radiation in the acute setting, and its potential adverse sequalae. We note a limitation was the absence of available testing for secondary resistance mutations that developed in this patient, which was highly susceptible to lorlatinib. However, current clinical guidelines (NCCN) do not specifically recommend resistance mutation testing upon progression, and outside of a clinical trial, the current algorithm supports lorlatinib as the agent of choice post-crizotinib failure.