Radiation damage is the major impediment for obtaining structural information from biological samples by using ionizing radiation such as x-rays or electrons. The knowledge of underlying processes especially at cryogenic temperatures is still fragmentary, and a consistent mechanism has not been found yet. By using a combination of single-crystal x-ray diffraction, small-angle scattering, and qualitative and quantitative radiolysis experiments, we show that hydrogen gas, formed inside the sample during irradiation, rather than intramolecular bond cleavage between non-hydrogen atoms, is mainly responsible for the loss of high-resolution information and contrast in diffraction experiments and microscopy. The experiments that are presented in this paper cover a temperature range between 5 and 160 K and reveal that the commonly used temperature in x-ray crystallography of 100 K is not optimal in terms of minimizing radiation damage and thereby increasing the structural information obtainable in a single experiment. At 50 K, specific radiation damage to disulfide bridges is reduced by a factor of 4 compared to 100 K, and samples can tolerate a factor of 2.6 and 3.9 higher dose, as judged by the increase of R free values of elastase and cubic insulin crystals, respectively.