The current study aims to determine the limit strains more accurately and reasonably when producing a forming limit curve (FLC) from experiments. The international standard ISO 12004-2 in its recent version (2008) states that the limit major strain should be determined by using the best-fit inverse second-order parabola through the experimental strain distribution. However, in cases where fracture does not occur at the center of the specimen, due to insufficient lubrication, the inverse parabola does not give a realistic fit because of its intrinsic symmetry in shape. In this study it is demonstrated that an inverse quartic function can give a much better fit than an inverse parabola in almost all FLC test samples showing asymmetric strain distributions. Using a quartic fit creates more reliable FLCs.