Background: The diagnosis of LM remains challenging, as standard diagnostic tools such as CSF cytology and MRI have low sensitivity. We previously showed that detection of CSF CTC by rare cell capture technology (RCCT) may be a more sensitive technique to detect LM from epithelial tumors at early stages. To validate those findings, we evaluated the performance of CSF CTC analysis in a new cohort of patients with a suspicion of LM. Methods: In this IRB-approved prospective study, patients with epithelial tumors who had clinical symptoms suspicious for LM were enrolled. All patients underwent MRI and CSF analysis through conventional cytology and enumeration of CSF CTC in that same sample.CSF CTCs were enumerated through an FDA-approved, EpCAM-based enrichment, utilizing RCCT immunomagnetic platforms and antibody covered ferroparticles (Janssen Diagnostics, Raritan, NJ). For the present analysis, samples were considered positive for CSF CTCs when there was at least one CSF CTC detected in a 3 mL sample ( ≥ 0.33 CSF CTC/mL). Diagnostic performance of CSF CTC was evaluated; the gold standard was either positive CSF cytology or unequivocal findings on MRI. Results: 62 patients were enrolled (27 breast, 20 lung, 7 GU, 3 GI, 2 renal and 3 other carcinomas). Among these patients, 21 (34%) had LM based on CSF cytology (n = 8), MRI findings (n = 4) or both (n = 9). CSF CTCs were positive in 27/62 samples (median 19.3 CSF CTC/mL, range 0.3 to 66.7), achieving a sensitivity of 95%, compared to 81% for CSF cytology alone and 62% for MRI alone. There were 7 false positive CSF CTCs, all of which with one or fewer CSF CTCs/mL; specificity was 83%. The one false negative sample also had negative CSF cytology but LM was seen on MRI. Conclusions: Detection of CSF CTCs by RCCT is a highly sensitive (95%) method to establish the diagnosis of LM, with superior diagnostic performance as compared to CSF cytology or MRI. However, results showing one or fewer CSF CTCs/mL are of uncertain significance and may correspond to false positive results. Larger studies with longer followup and ROC analysis are needed to investigate alternative CTC enumeration cut-offs to improve specificity.