PACS. 68.10.Gw -Interface activity, spreading. PACS. 68.15.+e -Liquid thin films. PACS. 68.35.Rh -Phase transitions and critical phenomena.Abstract. -We present the first microscopic images of the prewetting transition of a liquid film on a solid surface. Pictures of the local coverage map of a helium film on a cesium metal surface are taken while the temperature is raised through the transition. The film edge is found to advance at constant temperature by successive avalanches in a creep motion with a macroscopic correlation length. The creep velocity varies strongly in a narrow temperature range. The retreat motion is obtained only at much lower temperature, conforming to the strong hysteresis observed for prewetting transition on a disordered surface. Prewetting transition on such disordered surfaces appears to give rise to dynamical phenomena similar to what is observed for domain wall motions in 2D magnets.Introduction. -Over the past twenty years, prewetting transition has attracted a lasting interest since its first consideration by Ebner, . Considering an equilibrium between two fluid phases, the prewetting transition on a surface is simply the continuation off coexistence of the wetting transition. The surface coverage corresponding to the wetted state is simply finite rather than infinite. Apart from very special matching conditions [4,5] for which critical wetting transition has indeed been observed [6], first order transitions are expected, resulting from long range Van der Waals forces between the surface and the various fluid components [2][3][4]. In the simple case of a liquid/vapor phase equilibrium in presence of a surface, the surface coverage changes from a thin coverage to a thick liquid film. Experimental observations of these transitions on various systems have been made over the last decade [7][8][9][10][11][12][13][14][15][16][17][18][19][20][21][22]. A particular goal of experimentalists has been to check the first order nature of the transition, to determine the prewetting line in the T -∆µ plane and to locate the prewetting critical point. As a matter of fact, experiments have raised further questions about hysteresis phenomena, about the consequences of surface disorder, about the nucleation of the new phase at the transition and its subsequent growth. Presently there is no unique interpretation of the various experimental observations of prewetting transitions, especially at low temperatures. To our best knowledge, nobody has ever provided a clear