The cold, dense plasma in the polar magnetosphere of Ganymede has been observed to contain one species of positive ions. The usual assumption that they are protons leads to severe problems, e.g. finding an adequate ionization mechanism for hydrogen, or explaining why the expected oxygen ions are not found and what happens to the oxygen. We propose that the observed ions are in fact singly charged oxygen ions rather than protons, which immediately removes all these problems. We show that the assumption of O+ instead of H+ is fully consistent with the plasma ion measurements and, by multiplying the inferred number density by 4 and speed by 1/4, brings them into better agreement with independent estimates of electron density and ion bulk flow. ¿ 2000 American Geophysical Union