We clinically evaluated three commercially available (1→3)-β-D-glucan measurement methods for the diagnosis of fungal infections: the kinetic turbidimetric, the end-point chromogenic, and the kinetic chromogenic Limulus tests. For the kinetic turbidimetric assay, we examined a dilution and heating pretreatment method at 70°C for plasma to eradicate irregular (1→3)-β-D-glucan values caused by changes in the nonspecific turbidity of the reaction mixture, a problem occasionally encountered when the ordinary method is used (pretreated at 80°C). With the 70°C-pretreatment method, the nonspecific turbidity was eradicated and the values of (1→3)-β-D-glucan were lower than those obtained by the ordinary method. Thus, the 70°C-pretreatment method was less sensitive than the ordinary method, but its specificity was higher. The overall agreement rates between the kinetic turbidimetric assay and chromogenic assay were near 90%. Some samples in the control group showed unexpectedly high (1→3)-β-D-glucan values with the kinetic chromogenic assay. The measurement with the end-point chromogenic assay was influenced by sulfa drugs. In the kinetic turbidimetric method, the 70°C-pretreatment method is more effective than the ordinary method.