The guarded parallel-plate technique was employed on a black coal sample for an accurate measurement of the thermal conductivity over the temperature range from 298 to 496 K. The combined expanded uncertainties of the temperature (T) and thermal-conductivity (lambda) measurements at the 95% confidence level with a coverage factor of k = 2 are estimated to be 20 mK and 5%, respectively. It was experimentally observed that the measured thermal conductivity (lambda) of the wet and dry coal samples increases with temperature passes through a maximum around 390 K, and then it decreases gradually at higher temperatures. We attribute this maximum to the evolution of the volatile matter (VM) (devolatilazation) and aromatization of the carbon (pyrolysis), which is known to occur under heat treatment, and therefore, tends to increase the thermal conductivity. Over the experimental temperature range, the measured thermal-conductivity varied from 0.341 to 0.497 W.m(-1).K-1 for wet coal samples before thermal treatment and from 0.272 to 0.316 Wm(-1).K-1 for dry samples after thermal treatment. A considerable difference in thermal conductivity behavior was observed for the primary (originally nonthermally treated) and the second (thermally treated) repeated run. No temperature maximum or minimum (regular behavior) was observed in the thermal conductivity behavior for the thermally treated coal sample. The observed temperature behavior of the black coal's thermal conductivity (lambda) is a result of the complexity of the temperature behaviors of a, C-p, and rho, i.e., is the superposition of various temperature behaviors of a, C-p, and rho, and it reflects the temperature behavior of the heat capacity. This means that the temperature behavior of C-p dominates the thermal diffusivity in lambda = rho aC(p). This means that the temperature behavior of lambda and C-p is strongly correlated.