Major steps in our knowledge of high redshift galaxies and quasars were made possible by the synergy between the Hubble Space Telescope (HST) and 8–10 m class optical-near infrared ground-based telescopes. The galaxy samples were built from very deep, pencil-beam surveys with the HST and deep, wider area surveys undertaken with ground-based facilities. The most efficient selection technique of high-redshift star-forming galaxies is based on the presence of the Lyman-break within a given broad-band filter. Spectroscopic confirmation is then achieved with 8–10 m class telescopes. The first comparison of the galaxy Luminosity Function (LF) between two cosmic epochs was made for Lyman-Break Galaxies (LBGs) at z ∼ 3 and z ∼ 4 [1]. At these bright magnitudes, there is little evidence for an evolution of the LF shape and star formation density between these two epochs, whereas this is no longer the case at higher z (mostly photometric samples at z 5). There is a deficiency of luminous galaxies at z∼ 6, with a brightening of M by ∼ 0.6 magnitude between z∼ 6 and z∼ 3, but little evolution of the luminosity density and Star Formation (SF) rate density [2]. Sizes of high z galaxies have been measured from HST deep images [2, 3]. They are small: the mean half-light radius is equal to rhl = 2.3((1 + z)/3)−1.05 kpc (assuming Ω , Ωm, h = 0.7, 0.3, 0.7). This corresponds to only a small decrease in apparent size between z = 3 and z = 6, from rhl = 220 mas to 175 mas. At z ∼ 2, the brightest galaxies are several times larger. The surface density, n, of high redshift LBGs is derived from deep HST surveys. Bright LBGs at z ∼ 6, i-dropouts with z884,AB ≤ 25.4, are rare with n 0.015 arcmin−2. The surface density of idropouts increases rapidly with decreasing luminosities with n (z884,AB ≤ 28.1) 1.4 arcmin−2 [4]. At z∼ 7–8, one expects about a few tenths of galaxies per arcmin2 at these faint AB magnitudes. Another powerful technique to discover high-redshift galaxies is the detection of Lyα emission from narrow-band imaging. These surveys uncover a star-forming, young galaxy population, while the Lyman-break technique selects both young and older age galaxies. There are several hundreds candidate Lyα Emitters (LAEs) at z > 5.5. They are confirmed by subsequent spectroscopy. Their observed equivalent widths are large, wobs(Lyα) ≥ 120 A. The number density of LAEs at z ∼ 5.7 is