Reinforced concrete structures usually show cracks that are produced by environmental or stress conditions. Even if they do not represent a failure up to specific regulated criteria, they promote degradation thus reducing their service life. To prevent this occurring, continuous monitoring, maintenance and repairing is mandatory. The process not only consumes budget, but it is difficult, as many of these structures are in continuous service and accessibility of certain areas is challenging. Although concrete possesses autogenous self-healing capabilities, their effectiveness depends on many variables as the presence of water, Ca++, shape of the cracks, age of the concrete. Thus, in recent years, methods for repairing the cracks by autonomic self-healing have been proposed. Among them, the use of mineral crystalline admixtures is the most promising. They produce crystalline or expansive products in response to a crack. Nevertheless, a limited literature on this subject exists. The present work aims to investigate the automatic self-healing of two types of concrete, manufactured in accordance with UNI 11104 and in exposure class XC3 and XS1 respectively, containing two different self-healing additives. The mix-designs without them were used as reference. Mechanical properties of the different concrete mixtures were determined. Several specimens of each mixture were also subjected to compressive loads at load levels equal to 85% of their respective ultimate strength and then immersed in water for 60 days to investigate the self-healing properties. Cracks before and after the treatment were analysed. Finally, the specimens were tested again in compression. Ultrasonic procedures capable of estimating the linear and nonlinear response of the samples during the compressive tests were developed for the non-destructive evaluation (NDE) and structural health monitoring (SHM) of concrete. Results showed that mineral crystalline admixtures promote sealing of small cracks and improve the residual compressive strength when CEM II A-LL is used.