Geotechnical site characterization requires the determination of the stratigraphic profile, the groundwater level position, and the appropriate geomechanical design parameters required for each project. It is also necessary to assess both the spatial and temporal site variability. The shear wave velocity profile (Vs) and, consequently, the maximum shear modulus (G₀) are very important parameters for geotechnical projects. Although a large part of Brazil is covered by tropical soils, the literature on the dynamic behavior of these soils is limited. The seismic dilatometer test (SDMT) has been used in several site investigation campaigns, both at research test sites and in construction projects. This paper presents comparisons of the maximum shear modulus (G₀) profiles at two well-documented tropical soil sites using the SDMT and other in situ testing techniques such as the crosshole (CH) and downhole (DH) methods, the seismic piezocone penetration test (SCPTu), and seismic standard penetration tests (SSPTs). G₀ measurements obtained with the SDMT were found to be generally accurate and in good agreement with the CH, DH, SCPTu, and SSPT data used as references in the comparisons. Furthermore, this paper shows that the classical empirical correlations used to estimate G₀ from the intermediate parameters (E_D, K_D, and M_DMT) of the flat dilatometer test (DMT) were not able to estimate G₀ for both studied sites. This may be related to the presence of microstructure (e.g., cementation/bonding and aging) as well as a possible soil suction influence typical of the investigated tropical soil sites. The G₀/E_D and G₀/M_DMT ratios determined by the SDMT were used to assess the presence of microstructure and any possible soil suction influence. In addition, an equation was proposed to estimate G₀ from the intermediate DMT parameters.