In this paper the dynamic behavior of an oscillating tip-microlever system at the proximity of a surface is discussed. We show that the nonlinear behavior of the oscillator is able to explain the high sensitivity of the oscillating tip microlever and the observed shifts of the resonance frequency as a function of the tip surface distance without the need of introducing a particular short range force. PUBLISHED IN Phys. Rev. Lett. 82(17), 3388-3391 (1999).PACS numbers: 07.79. Lh, 61.16.Ch, 68.35.Bs A few years ago, the resonant noncontact mode (hereafter noted NC-AFM) was developed to map tip surface interaction as shifts of the resonance frequency [1,2]. Since then, numerous experimental efforts have been performed showing that measurements of shifts were able to produce images at the atomic scale [3][4][5][6][7]. The "routine" achievement of the atomic resolution with the NC-AFM [7] is thought of as a real breakthrough in the field of the scanning probe microscopy. Moreover, the use of a high vibrating amplitude of the cantilever (CL) with the NC-AFM and the development of the intermittent contact (the so-called tapping mode) [8][9][10] have boosted theoretical works dedicated to a nonlinear analysis [11][12][13][14].Before going a step further, it is worth discussing general ideas at a qualitative level. Leaving aside the technical point that the use of a high amplitude reduces the influence of the fluctuations in frequency [2], a first, counterintuitive, result is the use of a large vibrating amplitude allowing the atomic resolution to be achieved. Typically, the amplitudes used are 100 times greater than the vertical motion of the surface required, maintaining a given value of the frequency shift. The most common and shared idea is that to probe an attractive field with an oscillator requires the use of a small amplitude in order to keep the force field nearly constant throughout the vibrating amplitude. The idea is that shifts in resonance frequency are uniquely due to the gradient force variations.An image with the atomic resolution based on a repulsive interaction appears quite easily understandable, but getting the same resolution when the attractive van der Waals forces are involved is very puzzling. The van der Waals atom-atom interaction and the finite size of the interacting objects lead to a tip-sample interaction smoothly varying as a function of the CL-surface distance. For instance, the sphere-plan interaction leads to an attractive force with a d 22 power law. The force gradient dependence, with a power law d 23 , gives a too smooth variation of the resonance frequency and is unable to predict the observed shape of the frequency changes as a function of the tip-sample distance.