We present a multi-wavelength study of galaxy populations in the core of the massive, X-ray luminous cluster XMMU J2235 at z = 1.39, based on high quality VLT and HST photometry at optical and near-infrared wavelengths. We derive luminosity functions in the z, H, and Ks bands, approximately corresponding to restframe U, R and z band. These show a faint-end slope consistent with being flat, and a characteristic magnitude M∗ close to passive evolution predictions of M∗ of local massive clusters, with a formation redshift z > 2. The color − magnitude and color-mass diagrams show evidence of a tight red sequence (intrinsic scatter  ≲ 0.08) of massive galaxies already in place, with overall old stellar populations and generally early-type morphology. Beside the red colors, these massive (>6 × 1010   M⊙) galaxies typically show early-type spectral features, and rest-frame far-UV emission consistent with very low star formation rates (SFR < 0.2   M⊙ yr-1). Star forming spectroscopic members, with SFR of up to  ~100 M⊙/yr, are all located at clustercentric distances  ≳ 250 kpc, with the central cluster region already appearing effectively quenched. Most part of the cluster galaxies more massive than 6 × 1010   M⊙ within the studied area do not appear to host significant levels of star formation. The high-mass end of the galaxy populations in the core of this cluster appears to be in a very advanced evolutionary stage, not only in terms of formation of the stellar populations, but also of the assembly of the stellar mass. The high-mass end of the galaxy stellar mass function is essentially already in place. The stellar mass fraction estimated within r500 (~1%, Kroupa IMF) is already similar to that of local massive clusters. On the other hand, surface brightness distribution modeling of the massive red sequence galaxies may suggest that their size is often smaller than expected based on the local stellar mass vs. size relation. An evolution of the stellar mass vs. size relation might imply that, in spite of the overall early assembly of these sources, their evolution is not complete, and processes like minor (and likely dry) merging might still shape the structural properties of these objects to resemble those of their local counterparts, without substantially affecting their stellar mass or host stellar populations. Nonetheless, a definite conclusion on the actual relevance of size evolution for the studied early-type sample is precluded by possible systematics and biases.