We evaluated the diagnostic performance of the β-d-glucan (BDG) test (Beijing Gold Mountain River Tech) in diagnosing invasive fungal disease (IFD) and its variations among patients with different risks. Patients≥18 years old who underwent a serum BDG test (positive cutoff value>80 pg/ml) from April 2017 through May 2018 were collected consecutively. Patients were classified into three groups: group 1, patients with host factors as defined by the prior 2008 European Organization for Research and Treatment (EORTC) criteria; group 2, those with extended host factors in 2020 EORTC criteria; and group 3, those without any risk factor mentioned in the criteria. IFD was defined by 2020 EORTC criteria, but BDG was not considered. Diagnostic performance of the serum BDG test was measured by the area under the curve (AUC) of the receiver-operating characteristic curve. Among 469 patients, 15.4% (72/469) were diagnosed with IFD (48/191 [25.1%], 14/144 [9.7%], and 10/134 [7.5%] in groups 1, 2, and 3, respectively). The BDG assay showed fair performance (AUC 0.748 [95% CI: 0.688-0.810]). Sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) were 77.8%, 60.7%, 26.4%, and 93.8%, respectively. PPV was higher in group 1, and NPV was higher in group 3. Additionally, diagnostic odds ratios were 6.73, 2.88, and 5.92 in groups 1, 2, and 3. Immunosuppressant use, non-IFD/Candida colonization, and central venous catheter were associated with false positivity. Clinicians should cautiously interpret the BDG assay, considering the various diagnostic performances depending on the different levels of risk.