In-situ concrete strength assessment is commonly based on the implementation of Non-Destructive Tests (NDT) and coring of several specimens for Destructive Tests (DT). The most widely used non-destructive (ND) techniques are rebound hammer and ultrasonic pulse velocity. From the resulting data of NDT and compressive strength test results on specimens, a conversion law is identified which can then be used to estimate the compressive strength at any location in the structure. Since this process is very sensitive to many uncertainty sources, recent RILEM recommendations have revised the assessment paradigm by considering it in a probabilistic framework. The challenge is no longer to estimate the true concrete strength, but to limit the risk of a wrong estimation within a certain admissible interval. In this context, it is shown how synthetic simulations are used to capture the main features of this problem and are a firm basis for justifying the relationship between the most influencing parameters and the quality of the assessment. In addition, the effect of conditional coring, which consists in the selection of coring positions on the basis of previous NDT results, is highlighted and quantified. Finally, the methodology that makes possible to define the required number of cores for a given target accuracy is presented. © 2020 Elsevier Ltd

cited By 0