Hydrogen is widely in use in rocket propulsion systems, and as such, leakage of hydrogen from high-pressure fuel tanks requires accurate quantification. Safety concerns have led to the practice of conducting leak tests with helium (an inert gas) and to try to infer the hydrogen leak rates from helium data, often employing assumptions of essentially isentropic flow processes and choked leak orifices. The experimental study sought to quantify precisely the relationships between hydrogen and helium leak rates for various types of leaks. Simulated leak sources were fabricated by micromachining leaks or holes of prescribed shapes and cross-sectional areas in silicon wafers, utilizing the processes of photolithography and deep reactive ion etching. Dual thermal conductivity detectors were used to evaluate helium and hydrogen leak rates and to quantify differences in discharge coefficients among the various microorifices. Based on this quantification, the standard helium signature test procedure was found to underpredict hydrogen leak rates, in some cases significantly, if the corresponding helium tests are conducted at much lower pressures than those at which hydrogen leak rates are sought.