The prime directive in diagnosing neuroinfectious disease is to answer 3 deceptively simple questions: (1) Is the patient infected? (2) If so, with what? (3) How do we treat the infection? Our ability to answer the first 2 questions has been substantially improved by the emergence of multiplex approaches to diagnosis of central nervous system (CNS) infections, as exemplified by metagenomic next-generation sequencing (mNGS). Eight years ago, the first publications of proof-of-concept case reports were just appearing, establishing that mNGS could lead to actionable diagnosis with clear treatment implications, as in a case involving neuroleptospirosis in an immunocompromised child.1 The passage of time has enabled a more mature assessment of the role and limitations of mNGS in neurodiagnostics. The prospective Precision Diagnosis of Acute Infectious Diseases (PDAID) Study2 enrolled 204 patients with idiopathic encephalitis (64%), meningitis (34%), or myelitis (2%) from 8 hospitals. Of the 58 CNS infections ultimately diagnosed, 13 (22%) were only diagnosed using mNGS, including cases of St Louis encephalitis virus, hepatitis E virus, and Streptococcus agalactiae; 8 of these diagnoses affected clinical care. Overall, mNGS was highly concordant with other cerebrospinal fluid (CSF) direct detection tests (ie, pathogen-specific polymerase chain reaction [PCR], culture, and antigen testing); nevertheless, 26 additional infections (45%) diagnosed with conventional testing were missed by mNGS.