Gas hydrate in deep-sea sediments is a potential energy resource, and has been the focus of extensive drilling research. However, direct evaluation of the amount of the gas hydrate in marine sediments has been difficult because the gas hydrate in recovered sediment cores is at least partly dissociated due to the drop in pressure and increase in temperature during onboard recovery. In this study, we apply a new method based on oxygen isotopic composition of the H2O fraction of both hydrate and mud sub-samples (delta(OH)-O-18 and delta(OM)-O-18, respectively) in order to evaluate the volume percentage of gas hydrate in core sections collected from the Japan Sea off Joetsu and Oki, which contain different fabrics of the hydrate within hemipelagic mud. We measured isotopic composition of CO2 equilibrated with H2O of the sub-samples of a small size (typically 0.3 cm(3)) carefully separated from the core sediments and sealed in glass vials. The volume percentage of gas hydrate (H in %) was determined using porosity of the mud sub-samples and oxygen isotopic composition of the bulk pore water (delta O-18(PW)) squeezed from a certain length of core sediment (typically 20 cm) including dissociated hydrate. 28 out of the 29 examined core sections indicate the relation in the isotopic values of the three components, delta(OH)-O-18 > delta O-18(PW) > delta(OM)-O-18, as expected from isotopic fractionation that enriches O-18 in the hydrate component. Evaluated H-values of the 28 sections ranged from 1.0% to 95.4% and, for most of the section, the H-value was clearly larger than the value estimated by the hydrate distribution on core images. Our new method can, in a simple manner, correct for the underestimation of hydrate amount caused as a result of dissociation during core handling. Our oxygen isotopic data of the hydrate and mud sub-samples fits poorly with the isotopic evolutional curve that assumes Rayleigh fractionation in a closed system. This implies that the pore water isotopic composition may have been homogenized by diffusion and advection of less O-18-depleted pore water from the surrounding sediments. Presence of micro-scale hydrate in the mud matrix was suspected for some sections from the Joetsu site, which present a small difference between delta(OH)-O-18 and delta(OM)-O-18 as well as high CH2/CO2 ratios in headspace gas. We suggest that this method, if carried out with careful and quick onboard sampling, is appropriate for the estimation of gas hydrate as an energy resource based on the amount of hydrate present in marine mud.