Abstract For general stress states the fracture locus, under strictly proportional loadings, may be viewed as a two dimensional surface. Deviations from that locus under non-proportional loadings have received limited attention in the literature. The general problem is quite complex because of unavoidably intertwined history effects on both damage-free plasticity and plasticity-induced damage. This work first reports on a systematic attempt at quantifying deviations from the proportional fracture locus given a fracture theory. Two generic types of nonproportional loading are considered, which are believed to represent a heuristic integrity basis for an infinity of possible loading histories. In doing so, emphasis is laid on that component of the deviation associated with damage, thus reducing the representation of plastic flow to its simplest form. The predictive capability of the simple theory is then employed to rationalize experimental trends from the literature.