The Hamaker constant is considered to be a unique value that should theoretically not change for a given material. However, experimentally determined Hamaker constants vary significantly. These changes are suspected to occur not because of experimental error, but because of actual variations in the Hamaker constant. The critical coagulation concentration of an aqueous slurry containing similar to 150-nm-sized dispersed hematite particles was obtained by conducting a dynamic light scattering investigation of the relationship between salinity and particle aggregation rate. Moreover, the Hamaker constant of hematite particles in water was derived using the Derjaguin-Landau-Verwey-Overbeek (DLVO) theory. The Hamaker constant significantly varied with pH and heat treatment of the particles (350 degrees C, 1 h); values of 4.2 x 10(-21), 7.8 x 10(-20), and 1.6 x 10(-19) J were obtained for the untreated (pH = 6 and pH = 10) and heat-treated particles (pH = 10), respectively. A few-nanometer-thick particle surface layer presumably causes these changes in the Hamaker constant. This layer mostly controls the van der Waals interactions, and its composition and structure differ from those of the particle interior. The layer is sensitive to external stimuli and can consequently be easily affected by measurement conditions.