Thermal shocks are applied to a 304L austenitic stainless steel plate with a pulsed laser. A stroboscopic reconstruction is used for infrared (IR) and visible light camera measurements. The displacement fields are measured with a digital image correlation (DIC) technique. Different IR devices are used to measure the temperature variations (i.e. medium wave camera and short wave pyrometry). Several ways of determining the emissivity or absorptivity are discussed. The complete 3D thermal loading is numerically determined by minimising the difference between experimental measurements and finite element analyses of thermal fields. An elastoplastic model is then used to compute mechanical fields that are compared with DIC measurements. 1. Introduction Thermal fatigue may occur in pipes of nuclear power plants due, for instance, to the turbulent mixing of two fluids that have different temperatures. To study the material and structure response to constrained temperature variations, several experimental setups have been designed in different laboratories. However the temperature measurements have only been punctual via thermocouples and out of the zone of interest to prevent crack initiation on the connexion with the sensor.[1–6] Based on these measurements , the mechanical equivalent strain variation in the crack initiation region is evaluated thanks to numerical thermomechanical simulations. The number of cycles to crack initiation under such an equivalent strain is then compared with the number of cycles to failure in classical isothermal uniaxial fatigue tests. In many cases, it appears that this number of cycles for crack initiation is lower in thermal fatigue, namely, that crack initiation predictions in thermal fatigue based on classical isothermal fatigue tests would be non-conservative.[5] Some mechanical hypotheses can be proposed to explain such discrepancies, but they will not be discussed herein. In this paper, an effort is made to better estimate the experimental variations of temperature and strain fields. A new testing setup is proposed where thermal shocks are applied with a pulsed laser beam while the thermal and kinematic fields on the specimen surface are respectively measured with infrared (IR) and visible light cameras.