Nowadays a vast majority of the steel produced worldwide is via the continuous casting process route because this is the most low-cost, efficient and high quality method to mass produce metal products in a variety of sizes and shapes. Most of the continuous casters are the initial manufacturing step of a product which is very close to the final shape, reducing the need for further finishing. During continuous casting the liquid steel is solidified under controlled conditions of heat extraction to a semi-finished product that can subsequently be processed until final shape is reached. However, there is no perfect process and cracking during solidification of continuously cast steel slabs has been one of the main problems in casting for many years. In literature many terms are used for the phenomenon of crack formation at temperatures close to the solidus temperature, e.g. hot tearing, hot shortness, hot cracking or solidification cracking. Regardless of the name, hot tears represent a failure that occurs during casting in the regions of a solidifying slab that are at temperatures between solidus (TS) and liquidus (TL) and are subjected to simultaneously acting tensile and compressive stresses. The three steel grades considered in this study are a Low carbon aluminium killed (LCAK) steel grade, a high strength low alloyed (HSLA) steel grade, micro-alloyed with extra additions of vanadium, nitrogen and niobium, and a low-range HSLA (LR-HSLA), with the same concentration of niobium but a lower concentration of vanadium and nitrogen. The purpose of this study has been to investigate the differences between these three steel grades, with respect to hot tearing sensitivity upon solidifying in a thin slab caster. This research originates from the industrial demand for defect-free high-speed casting of steels and this thesis focuses on the thermo-mechanical behaviour of steel grades in relation to the solidification conditions in the continuous casting mould. Although these steel grades are not very different in chemical composition, thermodynamic calculations showed that the two HSLA steel grades have different propensity to the peritectic reaction upon solidification due to the combination of elements in the chemical composition that are either ferrite or austenite stabilisers. In order to better understand the special behaviour of the HSLA grade, compared with the LR-HSLA and the potential precipitation of TiN, phase field microstructure simulations have been performed, showing that TiN can form already during the latest stage of solidification, even when very low amounts of this element are present. If TiN particles trigger the coalescence of dendrite trunks, then it is possible to understand why, for a given Ti content an increased N content present in the HSLA grade can help to reduce the risk of hot tearing. Segregating elements in steel can influence the hot tearing susceptibility as they can widen the brittle temperature range (?TB), displacing its lower limit to lower temperatures. Unique hot tensile tests were carried out in the temperature range involving solidification, where the results suggest that among the studied steel grades, the LCAK steel grade has the lowest strength during solidification and a broader ?TB. Fractographic studies of the samples fractures above the non-equilibrium solidus revealed low melting phases even at temperatures as low as 1360 ºC. These low-melting films do have an effect on the hot tearing behaviour of commercial steel grades. The LCAK is particularly prone to develop non-metallic particles at the last stages of solidification. Whilst the study with the Mould Cracking Simulator did not confirm the increased hot tearing susceptibility of the LCAK and LR-HSLA, it did partially substantiate that the modifications on its design are novel in the way that no other physical model of cracking during continuous casting is able to simulate the process in such a way that it resembles the industrial process. Despite the alloys tested during the first few experiments were not the same alloy compositions studied in this thesis, the results of the Mould Cracking Simulator are very promising. It is expected that this new physical model will help in the study of the thermomechanical properties of commercial steels motivated by the continuous trend towards improved casting productivity by increasing casting speed and higher quality, as well as for the development of new steel grades.