By employing inverse modeling to analyze the laboratory data, we determined the composite thermal conductivity (k(theta), W/m/K) of a porous methane hydrate sample ranged between 0.25 and 0.58 W/m/K as a function of density. The calculated composite thermal diffusivities of porous hydrate sample ranged between 2.59 x 10(-7) m(2)/s and 3.71 x 10(-7) m(2)/s. The laboratory study involved a large heterogeneous sample (composed of hydrate, water, and methane gas). The measurements were conducted isobarically at 4.98 MPa over a temperature range of 277.3-279.1 K. Pressure and temperature were monitored at multiple locations in the sample. X-ray computed tomography (CT) was used to visualize and quantify the density changes that occurred during hydrate formation from granular ice. CT images showed that methane hydrate formed from granular ice was heterogeneous and provided an estimate of the sample density variation in the radial direction. This facilitated quantifying the density effect on composite thermal conductivity. This study showed that the sample heterogeneity should be considered in thermal conductivity measurements of hydrate systems. Mixing models (i.e., arithmetic, harmonic, geometric mean, and square root models) were compared to the estimated composite thermal conductivity determined by inverse modeling. The results of the arithmetic mean model showed the best agreement with the estimated composite thermal conductivity.