The measurement of serum (1-3)-beta-D-glucan (beta-glucan) in cases with deep seated mycosis is a useful diagnostic method. Beta-glucan has usually been measured using two different methods: by an alkali treatment, chromogenic automated kinetic assay (chromogenic assay), and by detergent dilution and heating methods, kinetic turbidimetric assay (turbidimetric assay). However, there are often large discrepancies in the beta-glucan values measured by these two methods. In this study, we reexamined the values of beta-glucan obtained by the two techniques, using 343 serum samples from 146 patients who had been treated in Kawasaki Medical School between January 1999 and May 1999, and then analyzed the reasons for the differences. Serum beta-glucan results measured were evaluated by segregating them into three clinical categories: cases with proven deep mycosis, cases with probable deep mycosis and cases without deep mycosis. In addition, the beta-glucan in the samples was suppressed by carboxy-methylated curdlan (CM-curdlan), and then was remeasured to find a non-specific reaction. Although a certain correlation was found between the serum beta-glucan results measured by the two methods, the values measured by the chromogenic assay were, in general, higher than those measured by the turbidimetric assay. There were also many samples in the cases without deep mycosis that showed positive values with the chromogenic assay, but not with the turbidimetric assay. With the turbidimetric assay, the addition of CM-curdlan suppressed the values of beta-glucan in all samples; however, when measured by the chromogenic assay the values in many samples remained high. These results suggest that a non-specific reaction which did not include beta-glucan was detected by the chromogenic assay. Further studies are needed to evaluate the characteristics and comparable usefulness of the two assays.