This paper addresses the issue of accurately describing the structures and properties of electronically excited systems embedded in an environment, through multiscale approaches combining quantum-mechanical (QM) and polarizable classical representations of the system and environment, respectively. Such approaches represent an efficient strategy and allow to effectively study the excited states of molecular systems in the condensed phase, still maintaining the computational efficiency and the physical reliability of the ground-state calculations. The most important theoretical and computational aspects of the coupling between the QM system and the polarizable environment are presented and discussed. Even if these approaches already reached an evident degree of maturity, they can still be subject to further development, in order to achieve their full potential. This perspective presents an overview of the state of the art of these strategies, showing the fields of applicability and indicating the current limitations, which need to be overcome in future developments. K E Y W O R D S excited states, linear response, polarizable continuum model, polarizable molecular mechanics, state specific 1 | INTRODUCTIONThe efficient exploration of the excited state (ES) energy surfaces of (supra) molecular systems is fundamental for the understanding of many photochemical and photophysical processes of technological and biological interest. Geometries, energies and properties such as dipole moments and transition and state densities of electronic ESs are the key ingredients to simulate, among the others, linear and nonlinear spectroscopic signals, energy and electron transfer processes, photocatalysis, light emission, and optical information storage. Most of these phenomena occur in condensed phase, where the molecular environment may largely affect the electronic structure reorganization following an electronic transition [1,2] Multiscale approaches [3,4] are some of the most powerful computational strategies for interpreting and simulating molecular processes in the condensed phase: an effective microsystem (which can be a single molecule or an aggregate of supramolecular size, henceforth system, or Sys) is assumed to be responsible for the observed process, property or signal, as its nuclear and electronic degrees of freedom (DoFs) are directly involved in it. On the other hand, the embedding macrosystem (henceforth environment, or Env) does not take part in the process, but rather acts as a perturbation, thus affecting the result of the measurement. The two components can mutually interact but they typically keep their number of nuclei and electrons constant during the process investigated (although so-called adaptive QM/molecular mechanics (MM) models exist to simulate processes where a dynamical definition of the QM/MM boundary is introduced [5,6] ). By applying this simplified but effective picture, a quite accurate and complete interpretation is generally achieved for systems of increasing complexity.