Abstract.A method to find the Excited-States Quantum Phase Transitions (ESQPT's) from paritysymmetry in the Dicke model is studied and presented. This method allows us to stablish a critical energy where ESQPT's take places, and divides the whole energy spectrum in two regions with different properties.Keywords: Quantum phase transition, Spontaneous symmetry-breaking PACS: 42.50.Nn, 05.30.Rt, 11.30.Qc, 64.70.Tg The Dicke model describes the interaction between an ensemble of two-level atoms and a single electromagnetic field mode, as a function of the radiation-matter coupling [1]. Its main property is a second order Quantum Phase Transition (QPT) which produces a macroscopic population of the upper atomic level. The Hamiltonian can be written as follows:where the N atoms are represented by J, the angular momentum operator with a pseudospin lenght J = N/2. Photons are represented by the usual annihilation and creation operators, and ω and ω 0 represent the frecuency of the cavity mode and the transition frecuency respectively. The system is governed by λ , the intensity of radiation-matter coupling, and its critical value for QPT is λ c = √ ωω 0 /2. We consider thath = 1 and ω = ω 0 = 1. The parity of the system is Π = e iπ(J+J z +a † a) , which is a conserved quantity since [H, Π] = 0. The symmetry-parity is given by the invariance of H under J x → −J x and a → −a [2]. This symmetry is spontaneously broken when the critical point of QPT is crossed [3]. Since the parity Π is a conserved quantity, we can label the eigenstates of H with a certain value of the parity, positive or negative. Hence, we obtain two sets of eigenstates, for a i-th eigenstate we have another with opposite parity. Fixing he number of atoms N, we analyze the relative difference of the energy between two i-th eigenstates with opposite parity. We have shown that for λ > λ c every couple of eigenstates with opposite parity are degenerated below a certain critical energy [4]. This critical energy E N c (λ ) separates two regions in the spectrum and is a function of λ and the number of atoms N. Here we show how to find the E c in the thermodinamical limit by fitting the finite precursor E N c (λ ) to a linear function E N c (λ )/J = A N + B N λ , and studying the finite-size scaling of coefficients A N and B N .