Background: While circulating tumor DNA (ctDNA) is a routine tool for therapeutics in metastatic solid tumors, data in those with stable and active brain metastases (BrMs) is limited. We assessed genomic alterations in ctDNA from patients (pts) with solid tumor BrMs in three groups: isolated BrMs/stable extracranial disease (iCNS), concurrent brain and extracranial progression (cCNS), and extracranial progression with no active BrMs at ctDNA draw (eCNS). We also compared ctDNA alterations between pts with and without BrMs. Methods: Pts with BrMs and a Guardant 360 profile (n=253) from the Duke Molecular Registry of Tumors between 01/2014 – 12/2020 were identified. Intracranial and extracranial disease status were determined via investigator assessment (RECIST and RANO-BM criteria) within 30 days of ctDNA test. Mutant allele fraction (MAF) was defined as the % of mutant alleles divided by total alleles at a given loci. Differences in mean ranks of the MAF % maximum were compared using the Kruskal-Wallis test and pairwise comparisons with the Dwass, Steel, Critchlow-Fligner multiple comparisons post-hoc procedure. We compared differences in MAF >1% using the chi-square test. Results: Among the 253 pts with BrMs: 29 (12%) had iCNS, 160 (63%) cCNS, and 64 (25%) eCNS. The two most common tumor types were breast (BC, 12.0%) and non-small cell lung cancer (NSCLC, 76.4%). Median MAF was lower in the iCNS group (1.6%) vs cCNS (4.8%) p=0.002, but eCNS did not differ (2.6%). In pts with BC BrMs, ESR1 mutations were most frequent in the iCNS group (67%), followed cCNS (54%) then eCNS (18%) (p = 0.09) as were PIK3CA mutations, iCNS(50%), cCNS (46%), and eCNS (27%), p = 0.55. In pts with NSCLC BrMs, EGFR mutations were most frequent in the iCNS followed by cCNS then eCNS (67%, 40% and 37%, p = 0.08). KRAS mutations were more frequent in cCNS followed by eCNS and iCNS (30%, 17% and 6% p = 0.031). Patients with no BrMs (n=449) included 24% breast and 76% NSCLC. In the ctDNA, ESR1 and BRCA2 mutations were more frequent in pts with BC BrMs than those without (60% vs 25%, p< 0.001) and (17% vs 5%, p=0.022), respectively. Mutations in EGFR were more frequent in pts with BrMs vs without (36% vs 17%, p<0.001). Sequencing from both BrMs tissue and ctDNA were available for 8 pts (4cCNS and 4 iCNS). 7 of these (87.5%) had identical mutations in tumor and ctDNA. Conclusions: This study illustrates the feasibility of detecting actionable mutations from ctDNA among pts with BrMs, with and without extracranial disease. A higher frequency of actionable mutations in several targetable genes was observed in ctDNA when comparing pts with and without BrMs, thus providing therapeutic opportunities. Additional studies comparing ctDNA and alterations in BrMs tissue are needed to determine if ctDNA can be considered a surrogate to support treatment decisions.