BACKGROUND: Metagenomic next-generation sequencing (mNGS) was suggested to potentially replace traditional microbiological methodology because of its comprehensiveness. However, clinical experience with application of the test is relatively limited. METHODS: From April 2017 to December 2017, 511 specimens were collected, and their retrospective diagnoses were classified into infectious disease (347 [67.9%]), noninfectious disease (119 [23.3%]), and unknown cases (45 [8.8%]). The diagnostic performance of pathogens was compared between mNGS and culture. The effect of antibiotic exposure on detection rate was also assessed. RESULTS: The sensitivity and specificity of mNGS for diagnosing infectious disease were 50.7% and 85.7%, respectively, and these values outperformed those of culture, especially for Mycobacterium tuberculosis (odds ratio [OR], 4 [95% confidence interval {CI}, 1.7-10.8]; P < .01), viruses (mNGS only; P < .01), anaerobes (OR, ∞ [95% CI, 1.71-∞]; P < .01) and fungi (OR, 4.0 [95% CI, 1.6-10.3]; P < .01). Importantly, for mNGS-positive cases where the conventional method was inconclusive, 43 (61%) cases led to diagnosis modification, and 41 (58%) cases were not covered by empirical antibiotics. For cases where viruses were identified, broad-spectrum antibiotics were commonly administered (14/27), and 10 of 27 of these cases were suspected to be inappropriate. Interestingly, the sensitivity of mNGS was superior to that of culture (52.5% vs 34.2%; P < .01) in cases with, but not without, antibiotic exposure. CONCLUSIONS: mNGS could yield a higher sensitivity for pathogen identification and is less affected by prior antibiotic exposure, thereby emerging as a promising technology for detecting infectious diseases.