The Fourier transform of the observed magnetic quantum oscillations (MQO) in YBa2Cu3O 6+δ high-temperature superconductors has a prominent low-frequency peak with two smaller neighbouring peaks. The separation and positions of these three peaks are almost independent of doping. This pattern has been explained previously by rather special, exquisitely detailed, Fermi-surface reconstruction. We propose that these MQO have a different origin, and their frequencies are related to the bilayer and inter-bilayer electron hopping rather than directly to the areas of tiny Fermi-surface pockets. Such so-called "slow oscillations" explain more naturally many features of the observed oscillations and allow us to estimate the inter-layer transfer integrals and in-plane Fermi momentum.Magnetic quantum oscillations (MQO) provide a traditional and powerful tool to study the Fermi surface and other electronic structure parameters of various metals. [1][2][3] ([4, 14-21] reviewed in [5,[22][23][24]). The Fourier transform of these quantum oscillations has a prominent peak at frequency F α ≈ 530T with two smaller shoulders at F ± = F α ± ∆F α , where ∆F α ≈ 90T . All these frequencies are much smaller than expected from closed pockets of the Fermi surface(FS). Many different theoretical models have been proposed to explain such a set of frequencies ([25-29], reviewed in [5,22,23]). While these interpretations vary in detail, such as inclusion of spin-orbit or Zeeman splittings, they are all based on Fermi-surface reconstruction due to the periodic potential created by a charge density wave (CDW). A weak, probably, inhomogeneous or fluctuating CDW order has been detected in YBa 2 Cu 3 O 6+δ compounds by X-Ray scattering [30][31][32][33], nuclear magnetic resonance [34,35], and sound velocity measurements [36]. High magnetic fields suppress superconductivity and lead to long-range CDW coherence.[33] Static CDW order can indeed lead to Fermi surface reconstruction, seen in new MQO frequencies, but only if the CDW potential is sufficiently strong. More precisely, the CDW energy gap must be larger than the magnetic-breakdown gapwhere ω c is the cyclotron energy, i.e. the separation between the Landau levels, and E F ∼ 1eV is the Fermi energy of the unreconstructed electron dispersion. The oscillations in cuprates are measured in magnetic fields B higher than 30 tesla, where ∆ MB 40meV is rather large and a fluctuating CDW ordering may not be enough to form new frequencies with amplitudes sufficient for experimental observation. Note that a frequency pattern somewhat similar to that of YBa 2 Cu 3 O 6+δ is observed in the closely related stoichiometric compound YBa 2 Cu 4 O 8 , [37][38][39] where there is no experimental indication of a static superstructure. Even if this CDW is sufficiently strong, it is hard to explain the observed three-peak frequency pattern of MQO in YBCO without additional frequencies of similar amplitude from the CDW wave vector seen in X-ray experiments. [30][31][32][33] Moreover, if FS reconstruction really i...