Slender structures are usually characterized by large displacements and rotations but small strains. Many practical problems can be framed in this context. Some examples are metal and composite structures, often characterized by buckling phenomena and strong imperfection sensitivity. The standard approach to simulate the behavior of this kind of structures consists of the use of the finite element (FE) method, in order to transform the continuum problem into a discrete one. The nonlinear discrete equations, completed with an arc-length constraint defining the step size, are solved step-by-step by using the Newton iterative method, in order to evaluate the equilibrium path of the structure. Although this strategy is well-established, it can be plagued by its high computational cost, due to two different limitations: i) the number of FE degrees of freedom (DOFs) used to approximate the differential equations; ii) the number of iterations required to obtain an equilibrium point and then to trace the desired equilibrium path, once the continuum problem has been discretized. In this work attention is focused only on the second aspect. Most of the existing FE codes are based on displacement formulations, i.e. the kinematic field is interpolated and the discrete kinematic DOFs represent the unknowns of the non-linear equations. Other kinds of formulations are possible, like for instance the mixed (stress-displacement) one, also known as hybrid-stress, in which both the stress and the displacement fields are interpolated. When comparing mixed and displacement finite elements many authors observe that the mixed ones are more robust and allow larger steps in path following geometrically non-linear analyses. It was shown that the robustness and the efficiency (in terms of number of iterations) of the Newton iterative scheme are penalized in displacement formulations because of a phenomenon that was called ”extrapolation locking” [1]. This is not a locking of the FE discretization, but of the iterative scheme usually found in beam/shell problems, when the axial/membranal stiffness is much higher than the flexural one. In these cases the iteration matrix evaluated in the current estimates of the equilibrium point can be far from the optimal one, in terms of the convergence condition of the Newton method. The number of iterations required to evaluate the equilibrium path, quickly grows as the stiffness ratios of the structure increase while, at the same time, the step size required to avoid loss of convergence drastically decreases. This phenomenon, that is typical of FE displacement formulation, does not affect mixed FE formulations, which are free from ”extrapolation locking”. In recent years many other researchers have experienced the better behavior of the mixed FE models in geometrically non-linear analyses. A recent work [1] investigated the fast convergence and the high robustness of mixed FE models, providing a clear explanation of the origin of the phenomenon in a general context. It was shown that the evolution of the displacement iterative process is forced to satisfy the constitutive equations at each iteration and this constraint leads to a deterioration in the convergence properties. On the contrary in mixed FE models the stress DOFs are independent variables, directly extrapolated and corrected, which onlyat convergence satisfy the discrete constitutive equations. The mixed iterative scheme is then free to naturally evolve to the solution while the displacement iterative scheme is constrained to follow an assigned evolution in which the constitutive equations must be satisfied. Thus, the ready convergence of the Newton method for mixed FEs is not linked to the FE interpolation, but due to the different ”format” of the iterations. Mixed elements have a higher computational cost in constructing their element stiffness matrix and internal force vector with respect to displacement FEs, because of the stress extra-variables usually condensed at element level, but their use in geometrically non-linear problems is justified by their readier convergence. The question posed in this paper is ”is it possible to use a mixed iterative scheme without introducing a mixed FE interpolation?”. A positive answer to this question would have important implications since a great part of the existing code is based on well-established displacement FE interpolations. In this paper we will show that this is possible and a mixed format of the Newton method for geometrically non-linear structural problems discretized via displacement-based finite elements is presented. The strategy is inspired by the more efficient iterative scheme for mixed FE models. The idea consists of the relaxation of the constitutive equations at each integration point. In this way, the stiffness matrix of a displacement-based FE maintains its original form. The only difference is that the stresses at each integration point, used for the matrix evaluation, are directly extrapolated and corrected, i.e. used as independent variables. This leads to a ”better” iteration matrix, which allows a very low number of iterations and very large steps (increments) in step-by-step analyses. With respect to mixed FEs no stress interpolations are present, so avoiding any additional cost in the evaluation of the stiffness matrix. Furthermore the final equilibrium path is the same as the original displacement formulation since the constitutive law is recovered exactly at convergence. The method, that we call MIP (Mixed Integration Point) Newton, converges much faster than the standard Newton method, as shown by many numerical tests with different structural models and FEs. The gain in terms of number of iterations required is impressive as well as the very large steps that the MIP Newton can withstand without loss of convergence. The computational cost of a MIP iteration is the same as a standard one. Furthermore, the iteration matrix evaluated with the MIP strategy is so ”good” that the modified version of the method (MIP modified Newton), which computes and decomposes the iteration matrix at the first estimate of each equilibrium point, can be conveniently adopted. From the implementation point of view, a few changes to the standard approaches are required, without upsetting the existing FE codes and then its inclusion is straightforward. The performances of the proposal are tested by means of many numerical examples, regarding beams and shells discretized using finite elements [2] and isogeometric approaches [3]. In geometrically non-linear analysis, the proposed strategy is so robust, efficient and simple that it seems destined to replace the standard Newton method in any finite element code based on displacement formulations.