A method for assessing the susceptibility of a weld to transverse cracking is proposed based on a newly designed test specimen featuring a transverse gap in the centre to control restraint and produce a large residual stress in the weld metal. Through FEM simulation of the effects of the weld bead length on the longitudinal residual tensile stress (s x ) after welding, it is confirmed that the transverse gap can significantly increase s x after welding. Furthermore, with a fixed transverse gap length, s x increases monotonically with the length of the weld bead. Two different commercial weld wires were used to check the effectiveness of this method; the experimental results obtained indicate that the critical cracking length of the weld bead differs depending on the weld wire used. The critical length of the weld bead can therefore be used as an effective index of its susceptibility to transverse cracking.