The x-ray emission in laser-plasma accelerators can be a powerful tool to understand the physics of relativistic laser-plasma interaction. It is shown here that the mapping of betatron x-ray radiation can be obtained from the x-ray beam profile when an aperture mask is positioned just beyond the end of the emission region. The influence of the plasma density on the position and the longitudinal profile of the x-ray emission is investigated and compared to particle-in-cell simulations. The measurement of the x-ray emission position and length provides insight on the dynamics of the interaction, including the electron self-injection region, possible multiple injection, and the role of the electron beam driven wakefield. Remarkable advances in relativistic laser-plasma interaction using intense femtosecond lasers have led to the development of compact electron accelerators and x-ray sources with unique properties. These sources use the very high longitudinal electric field associated with plasma waves, excited in an under-dense plasma by a relativistic laser pulse, to trap and accelerate electrons to relativistic energies [1]. For laser and plasma parameters corresponding to the bubble or blowout regimes [2,3], the production of quasimonoenergetic electron beams was demonstrated [4]. During their acceleration, these electrons perform transverse (betatron) oscillations due to the transverse focusing force of the wakefields. This leads to the emission of bright and collimated femtosecond beams of x rays [5–7]. Such a compact and cost effective x-ray source could contribute to the development of emerging fields such as femtosecond x-ray imaging [8]. The x-ray emission can be exploited as well to provide information on the physics of laser-plasma interaction, such as electron tra-jectories in the bubble [9]. In this Letter, we demonstrate that by measuring the position and the longitudinal profile of the x-ray emission, one can determine important features of the interaction: laser pulse self-focusing, electron self-injection and possible multiple injection, as well as the role of the electron beam wakefield. The method relies on the observation, in the x-ray beam profile, of the shadow of an aperture mask adequately positioned just beyond the end of the emission region. The size of the shadow on the x-ray image permits us to determine the longitudinal position of the x-ray emission in the plasma, while the intensity gradient of the edge of the shadow yields the longitudinal profile of the emission. Because the x-ray emission position and length are closely connected to the electron injection position and the acceleration length, this measurement provides a unique insight into the interaction. Particle-in-cell (PIC) simulations are performed to analyze the experimental results. The experiment was conducted at Laboratoire d'Optique Appliquée with the ''Salle Jaune'' Ti:Sa laser system, which delivers 0.9 Joule on target with a full width at half maximum (FWHM) pulse duration of 35 fs and a linear polarization. The laser pulse was focused inside a capillary at 3:5 AE 1:5 mm from the entrance, with a f=18 spherical mirror. The FWHM focal spot size was 22 m, and using the measured intensity distribution in the focal plane we found a peak intensity of 3:2 Â 10 18 W Á cm À2 , corresponding to a normalized amplitude of a 0 ¼ 1:2. The target was a capillary made of two Sapphire plates with half-cylindrical grooves of diameter d cap ¼ 210 m and a length of 15 mm, filled with hydrogen gas at pressures ranging from 50 to 500 mbar. The target acts as a steady-state-flow gas cell [10]. The capillary wall surface roughness is around 1 m, and therefore x rays cannot be reflected by the capillary wall. In addition, from our laser contrast of 10 8 , f number and capillary diameter, we estimate that the pedestal intensity on the capillary wall is smaller than 10 7 W Á cm À2 , and thus no preplasma is formed before the x-ray pulse arrival. We measured either the x-ray beam profile using an x-ray CCD camera (2048 Â 2048 pixels, 13:5 Â 13:5 m 2), set at D ¼ 73:2 cm from the capillary exit and protected from the laser light by a 20 m Al filter, or the electron beam spectrum with a focusing-imaging spectrometer. In our experiment, the betatron emission had a divergence larger than the opening angle associated with the capillary exit, which acts as the aperture mask. The x-ray beam was thus clipped by the capillary [11]. Figures 1(a)–1(c) present a sample of shadows of diverse sizes corresponding to different longitudinal positions of the source, z X .