Hydrogen penetration into a metal leads to damages and mechanical degradations and its content measurement is of importance. For a precise measurement, a sample preparation procedure must be optimized through a series of studies on sample washing and drying. In this study, two-step washing with organic solvents and thermal soaking in inert gas were tried with a rod-shaped, API X65 steel sample. The samples were machined from a steel plate and then washed in acetone and etyl-alcohol for 5 minute each and dried with compressed air. After then, the samples were thermally soaked in a home-made nitrogen gas chamber during 10 minute at different heat gun temperatures from 100 to $400^{circ}C$ and corresponding temperature range in the soaking chamber was from 77 to $266^{circ}C$ according to the temperature calibration. Hydrogen residue in the samples was measured with a hot extraction system after each soaking step; hydrogen residue of $0.70{pm}0.12$ wppm after the thermal soaking at $77^{circ}C$ decayed with increase of the soaking temperature. By adopting the heat transfer model, decay behavior of the hydrogen residue was fitted into an exponential decay function of the soaking temperature. Saturated value or lower bound of the hydrogen residue was 0.36 wppm and chamber temperature required to lower the hydrogen residue about 95% of the lower bound was $360^{circ}C$. Furthermore, a thermal desorption spectroscopy was done for the fully soaked samples at $360^{circ}C$. Weak hydrogen peak was observed for whole temperature range and it means that hydrogen-related contaminants of the sample surface are steadily removed by heating. In addition, a broad peak found around $400^{circ}C$ means that parts of the hydrogen residue are irreversibly trapped in the steel microstructure.