<p><b>Abstract</b>—Measurements of the thermal conductivity give an effective value that includes the conductive thermal conductivity, convective component, and radiative contribution. To take into account the convection, there is the mathematical apparatus for calculation of the product of the Prandtl criterion and Grashof criterion from the physicochemical properties of a melt under study. A radiative contribution can be due to reemission of walls limiting a salt melt and also an immediate influence of the melt opacity in IR region. To take into account the latter, the data on optical properties of molten salts are necessary, but there are very few of them in literature, in particular, for fluoride systems. One of the methods of estimation and taking into account the effect of the radiative contribution to the thermal conductivity ensured by the “emissivity” of the melt is the “variable gap” technique. The stationary coaxial cylinder method is used to measure the temperature dependences of the thermal conductivity of molten mixtures FLiNaK and 0.85FLiNaK–0.15CeF₃ for various widths of the salt gaps (0.5, 1.0, 1.5, and 2.0 mm) using the “variable gap” technique. The contributions to the thermal conductivity of the salt melts of the convection and “emissivity” of the melts, and also reemission of the cylinder walls limiting the molten salt layer are evaluated. It is shown that the contributions of the convective and radiative components of the thermal conductivity of molten salts FLiNaK and 0.85FLiNaK–0.15CeF₃ are negligibly small, while the contribution of reemission of nickel surface walls reaches 5%.</p>