We consider individual decision-making where every alternative appears with a frame (a la Salant and Rubinstein (2008)). The decision maker is subject to inattention due to framing effects that leads to random choice. We characterize a frame-based stochastic choice rule according to which the choice probability of an alternative (say, x) is the probability with which attention is drawn by its frame and not by the frames which are associated with the alternatives that beat x according to a complete binary relation.