Previous data by our group demonstrated the antifungal efficacy of a vaccine consisting of laminarin (beta-(1,3)-glucan), conjugated with diphtheria toxoid, which generated protective anti-laminarin antibodies in mice. In this paper, we sought for the presence, isotype and subclass composition of natural anti-laminarin antibodies in an unselected population of human healthy subjects, in a comparison with antibodies directed against beta-(1,6)-glucan (pustulan) and branched beta-(1,3/1,6)-glucan (Pool 1) and mannan from Candida albicans. Almost all subjects showed detectable levels of anti-beta-glucan antibodies, with IgG largely prevailing on IgM, little, if any, IgA and no IgE. However, the titer of anti-beta-glucan antibodies was overall about 1log lower than that of anti-mannan antibodies of the corresponding isotype. In particular, the level of anti-laminarin IgG was the lowest one, its geometrical mean titer (95% confidence interval, CI) being 1838 (1245-2714) as compared to 8157 (6067-10,931) and 3940 (2911-5332) for pustulan and Pool 1 fungal glucan, respectively. Analysis of IgG subclass composition showed that IgG2 was the prevalent subclass against any antigen, and again the concentration of anti-laminarin IgG2 was the lowest one, its geometrical mean concentration being 0.13 (0.07-0.24)microg/ml as compared to anti-pustulan and anti-Pool 1 glucan and mannan IgG2 levels, which were 0.33 (0.2-0.5), 1.35 (0.9-2.0), and 36.1 (25.2-51.3)microg/ml, respectively. These data show that anti-laminarin antibodies are present at low levels in humans as compared to other anti-beta-glucan and, mostly, anti-mannan antibodies, and suggest that a protective antifungal vaccination in humans should attempt to tip the balance of antifungal antibodies in favour of the anti-laminarin ones.