For the first time, atomic oxygen column content [O]col has been inferred from June daytime monthly median foF1 and foF2 observations at Rome, Juliusruh, Sodankylä, and Boulder to analyze its long‐term variations for the period of ~6 solar cycles. The analysis is interesting in the light of possible anthropogenic impact on the upper atmosphere. After the removal of solar and geomagnetic activity effects from the inferred [O]col variations, the residual linear trends are negative and statistically insignificant at middle latitudes. It is shown that ~93% (the corresponding correlation coefficient is 0.964 ± 0.03) of the whole [O]col variability is explained by solar and geomagnetic activity long‐term variations and only ~7% may be attributed to other processes (reasons) including the anthropogenic impact. Solar and geomagnetic activity contributions to [O]col long‐term variations decrease with time, and this may be related to the low solar activity epoch, which we have entered. The main conclusion is that the long‐term variations of the atomic oxygen column content inferred from ionospheric observations are due to solar and geomagnetic activity; that is, they have a natural origin.