Adsorption, X-ray diffraction, and nuclear magnetic resonance analysis of subbituminous to high-volatile bituminous coal at in situ moisture content, low temperatures, and moderate pressures demonstrate that a significant proportion of the inherent moisture (nonmobile water) is available to form methane clathrate hydrates. These results have implications for coal gas resources and reservoir pressures in current areas of permafrost and the much larger regions that were glaciated during the Pleistocene. Methane adsorption tests indicate the clathrates form comparatively rapidly in coal micro- and mesopores from an immobile water phase, at lower pressures than those formed in macroporous materials. At successively higher experimental pressures, hydrates nucleate and grow rapidly on the scale of minutes to hours, until an “apparent” equilibrium pressure is reached. The onset of hydrate formation at 0 °C is at about 3.25 MPa for a Tarn coal with 33% moisture and at slightly higher pressures for the other coals with lower inherent moisture. The amount of gas consumed in hydrate formation, in excess of that attributed to sorption, is 11.6 cm3/g coal for the Tarn coal (Alaska) with 33% moisture, 8.15 cm3/g coal for a Dietz coal (Wyoming) with 22% moisture, and 1.85 cm3/g coal for a Texas coal with about 8% moisture. On a volume of gas (STP) per volume of water basis at 8 MPa, the Tarn and Dietz coal have similar values (35 and 37 cm3/cm3), whereas the Texas coal is measureably less (24 cm3/cm3). At 8 MPa and 0 °C, about 20% of the inherent moisture has participated in hydrate formation for the Tarn and Deitz coals and 13% for the Texas coal. If hydrate formation strips methane from the sorbed state, the proportion of water contributing to hydrate calculates was over 100% for the Texas coal, which suggests that methane is mainly forming from free gas. The methane hydrates analyzed by X-ray diffraction and nuclear magnetic resonance (NMR) spectroscopy are cubic (Pm3n̅ space group) and have small and large cage sI structures. NMR and PXRD spectra indicates that the small cages are about 90% occupied, while the large cages are completely full, yielding a stoichiometry of ca. CH4·6H2O, which is consistent with other natural hydrates. There is some evidence that some methane gas remains trapped in the smallest pores, whereas the hydrates occupy large pores, which may be due to the suppression of hydrate formation by high capillary pressures. The formation of methane hydrates, particularly in low-rank coals, markedly increases the capacity of the coal to store gas. Depending on the coal moisture content and coal rank, methane storage capacity in gas hydrates is up to 2 orders of magnitude greater than the gas storage capacity of the coal by sorption alone. Since low-rank coals invariably have high moisture content, if the strata lie with the hydrate stability zone, significant gas storage in hydrates is anticipated, if gas is available. At the low temperature required for hydrate formation, however, self-sourced methane from methanogenesis or thermal alteration is not anticipated. Successive formation and dissociation of methane hydrates during glacial and interglacial times in the Pleistocene can be anticipated to have impacted shallow gas reservoirs, including coals, to depth up to about 800 m, depending on the surface temperature and geothermal gradient. It is speculated that during the Pleistocene free and sorbed coal gas was scavenged during the formation of hydrates, and this process may explain the low gas content of some high-latitude coals. Similarly, lower than anticipated pore pressure and lack of free water in the Horseshoe Canyon coals in western Canada may be the result of dissociation of hydrates forming overpressures, which potentially could hydraulically fracture the coal and flush the coal of free water. Subsequent dissipation of the gas pressure would lead to the current low reservoir pressure.