Abstract
In view of a number of contrasting results for the cosmic-ray muon ratio, a careful calculation is carried out and extended down to Eμ=50 GeV. It is found that even with the large uncertainties involved the calculated μ+μ ratio is significantly higher than the experimental values over the energy range considered. Nuclear corrections have only a small effect if a coherent production model is used and cannot remove the discrepancy, which we conclude requires some new features of high-energy interactions, a higher np ratio in the primary flux, or both.