With the use of the Finite Element Method it has become possible to analyse and better understand complex physical processes such as the resistance welding by numerical simulation. However, simulation of resistance welding is a very complex matter due to the strong interaction between mechanical, thermal, electrical and metallurgical effects all signifcantly in uencing the process. Modelling is further complicated when down-scaling the process for welding micro components or when welding new advanced high strength steels in the automotive industry. The current project deals with three main themes aimed at improving the understanding of resistance welding for increasing the accuracy of numerical simulation of the process. Firstly methods for measuring and modelling mechanical and electrical properties at a wide range of temperatures is investigated, and especially the electrical contact resistance is addressed both theoretically and experimentally. Secondly the consequences of downscaling the process is investigated experimentally and discussed in relation to simulation of the process. Finally resistance welding of advanced high strength steels is addressed aimed at improving the simulation of the final weld properties. The temperature dependent material rheology of dierent advanced high strength steels and other materials, often resistance welded, were measured using hot tensile testing and hot compression testing. It is found that the Hollomon equation is capable of modelling material rheology at discrete temperatures with suffcient accuracy. Investigation of theoretical contact resistance models revealed that most models build on the classic theory by Greenwood and Holm. However, extensive simplifications and assumptions raise questions regarding the theoretical foundation of the models. Experimental measurements of contact resistance was performed on a Gleeble 1500 system, and the measurements revealed that surface hardness and film resistance interacts with the effect of pressure on the contact resistance. Numerical simulation of downscaled joints introduce problems not observed for large scale welding. Especially the relatively small electrode force and the formation of the actual contact area is found to cause discrepancies in simulation of micro spot welding, as well as in micro welding of joints of complex geometry where simulation of the collapse of the geometric projections presented problems. Simulation of two- and three sheet spot welding of advanced high strength steels DP600 and TRIP700 did generally agree well with experimental observations. Microstructure characterisation revealed that martensite was the main constituent in the final weld. By using empirical formulae by Blondeau et al. predicting martensite hardness, and a proposed average volume weighted material composition function, the predicted post weld hardness corresponded well with experimental observations.