In continuation of experiments by Glaeser and Wever wires of high purity Fe were quenched from temperatures up to 1430 °C with various quenching rates. The result was that with a carbon content of about 1 at. ppm the two recovery stages between 60 and 150 °C and between 180 and 300 °C which were observed earlier, do not exist. With a carbon content of about 10 at. ppm, however, the two stages are still observed even with quenching rates down to 10 K/s. This proves that these stages cannot be caused by the migration of vacancies. In slight variation of the interpretation of similar stages observed by Fujita and Damask in Fe with about 500 ppm C it is suggested that in the first stage carbon probably migrates as di‐interstitial to precipitations of ε‐carbide. These precipitations are dissolved again with rising temperatures beginning at 150 °C. The second stage is caused by competitive precipitation of Fe3C which dissolves also with rising temperatures between 300 and 500 °C. For quenching temperatures of 1100 and 1430 °C the height of the first stage increases with quenching rate. It is suggested that this is caused by dissolved oxygen which precipitated together with carbon within the temperature range of the first stage. For higher quenching rates dislocations are formed while crossing the α–γ‐transformation. These dislocations recover again between 800 and 400 °C. According to the results it is stated that the enthalpy of formation of vacancies is >1.65 eV and the enthalpy of migration is <1.0 eV.