Dissolution experiments of alumina and magnesia in silicate and aluminate melts were conducted at 1450, 1500 and 1550 °C in high temperature confocal laser scanning microscope (HT-CLSM) and continuous wear testing device (CWTD) to determine diffusivities. CLSM studies were performed with spherical sapphire and fused magnesia particles in three silicate and one calcium aluminate slags, whereas finger test of alumina and magnesia fine ceramics in CWTD were carried out in two silicate slags. Effect of rotational speed was also examined for alumina dissolution at 1550 °C in both slags. Laser measurement of CWTD includes the wear profile of whole sample surface with high resolution, as a result, the dissolution parameters calculated from these, are expected to be more accurate than the manual measurements of the post-mortem analysis. Three models were applied to determine diffusivity from CLSM studies. Shrinking core models with (M2) and without (M1) convective part of mass flux suffer from the fact they cannot represent the dissolution curve shape well, as their assumption regarding the effective diffusive boundary layer thickness proofs to be not accurate enough. Whereas the third model (M3) is more scientifically sound for quasi-steady state dissolution which incorporate the Stefan condition correctly, moving boundary condition and effect of bath movement. Dissolution time of the particle in CLSM increases with decreasing temperature, due to lower diffusivity. For the dynamic experiments of alumina fine ceramics with 200 rpm, Reynolds numbers were sufficiently high to suppress the Marangoni convection; however, this was not the case for magnesia dissolution. In these experiments, diffusivity was determined using Sherwood relations and simulation from dynamic corrosion experiments and the diffusivities show good agreement. Furthermore, these results are very similar to diffusivities of M3 from CLSM studies.