With significantly improved sample quality and instrumental resolution, we clearly identify in the (π,0) ARPES spectra from YBa2Cu3O6.993, in the superconducting state, the long-sought 'peakdip-hump' structure. This advance allows us to investigate the large a-b anisotropy of the in-plane electronic structure including, in particular, a 50% difference in the magnitude of the superconducting gap that scales with the energy position of the hump feature. This anisotropy, likely induced by the presence of the CuO chains, raises serious questions about attempts to quantitatively explain the YBa2Cu3O 7−δ data from various experiments using models based on a perfectly square lattice.PACS numbers: 74.25.Jb, 74.72.Bk, 79.60.Bm High-temperature superconductivity (HTSC) is intimately related to the CuO 2 plane, which is the only common structural feature in all cuprates. This fact has led most of the proposed microscopic theories to assume a CuO 2 square planar structure. However, for the practical reason of sample quality, some of the most important and defining experiments have been performed on YBa 2 Cu 3 O 7−δ (Y123), which does not have a square lattice, but rather an orthorhombic structure (b/a ≈ 1.015), caused by the presence of a CuO chain layer [1]. This orthorhombicity, according to LDA calculation [2], should result in significant anisotropy in the in-plane electronic structure (this term will be used throughout this paper to refer to the electronic states associated with the CuO 2 plane), making it problematic to compare theories based on a square lattice with experimental data from Y123. Therefore, it is crucial to quantify the effect of orthorhombicity, if any, on the in-plane electronic structure in Y123. The problem is that angle-resolved photoemission spectroscopy (ARPES), being a uniquely powerful tool for this important task, has, until now, not been particularly effective for the study of Y123 [3]. The important 'peak-dip-hump' structure, which is seen routinely in Bi 2 Sr 2 CaCu 2 O 8+δ (Bi2212) [4] has never been observed in Y123. This absence, together with the presence of a surface state [5], raises questions about ARPES data from Y123, and the universality of the superconducting peak in the cuprates.This paper reports a breakthrough in this important issue, made possible by significantly improved sample quality and instrumental resolution. By isolating a surface state peak near the Fermi energy (E F ), we can clearly resolve a 'peak-dip-hump' structure in the ARPES spectra around (π, 0) in Y123 that resembles the superconducting peak observed in Bi2212 [4]. More significantly, we find a strong a-b asymmetry of the in-plane electronic structure, such as the superconducting gap magnitude, which differs by about 50%. We argue that such a strong inplane a-b anisotropy should be taken into account when interpreting experiments performed on Y123.ARPES experiments were carried out at beamline 5-4 at SSRL, which is equipped with a normal-incidencemonochromator and a SCIENTA SES-200 analyzer. Untwinned Y...