Once again an attempt is made to quench‐in vacancies in very pure iron (< 0. 5 at ppm C). This is achieved with a quenching rate of 1.5 × 104 K/s and an iron wire with a diameter of 100 μm. The recovery begins at 380 K and continues up to stage IV. An annihilation in stage III cannot be observed. This result can be interpreted with vacancy–C pairs — formed above 380 K with the residual content of carbon — and a following dissociation of the pairs and an annihilation of the vacancies at sinks up to temperatures of 600 K. Such an interpretation assumes the validity of the “two interstitial model” (TIM) and a formation enthalpy of the V–C pairs of ≧4 eV. The possibility of precipitation of the residual carbon in the form of graphite is discussed.