In an earlier paper [R. G. Musket, D. W. Brown, and H. C. Hayden, Nucl. Instrum. Methods B 7/8, 31 (1985)] we showed that high‐dose oxygen ion implantation into bulk aluminum results in the formation of subsurface layers of Al2O3. However, for oxygen doses exceeding that required for oxidation of all the locally available aluminum atoms, the implanted layer contained excess oxygen with the ratio O/Al reaching 3.2. At that time we proposed that excess oxygen was trapped physically at point defects and/or microviods (i.e., bubbles). We have now examined these implanted layers using cross‐sectional transmission electron microscopy and cross‐sectional scanning electron microscopy. Both of these techniques have provided clear evidence for the existence of 20–100 nm oxygen bubbles. Combining these results with those from Rutherford backscattering and conventional transmission electron microscopy has led to a more complete understanding of the formation process.