We aimed to assess the accuracy of measuring serum or plasma (1→3)-β-D-glucan (BDG) for the diagnosis of invasive fungal infections (IFIs) by means of a meta-analysis of relevant studies. We searched in bibliographic databases for relevant cohort or case-control studies. We primarily compared BDG between patients with proven or probable IFIs (excluding Pneumocystis jirovecii infections), according to the criteria of the European Organization for Research and Treatment of Cancer/Mycoses Study Group or similar criteria, and patients without IFIs (excluding healthy individuals as controls). A total of 2979 patients (594 with proven or probable IFIs), included in 16 studies, were analyzed. The pooled sensitivity of BDG was 76.8% (95% confidence interval [CI], 67.1%-84.3%), and the specificity was 85.3% (95% CI, 79.6%-89.7%). The area under the summary receiver operating characteristic curve was 0.89. Marked statistical heterogeneity was noted. BDG has good diagnostic accuracy for distinguishing proven or probable IFIs from no IFIs. It can be useful in clinical practice, if implemented in the proper setting and interpreted after consideration of its limitations.