Framing effects are said to indicate irrationality in decision making because they illustrate that linguistically different descriptions of equivalent options lead to inconsistent choices. A review of the literature on the effects of adding, or subtracting, implicated complements of the sure option shows that this leads to a classic framing effect, a reversal of the classic effect, or no framing effect. Thus, the assumption of equivalence of formulations is not justified. In addition we provide a test of two major, but opposing theories on framing, prospect theory and fuzzy-trace theory. Based on an online study we investigated the effects of subtracting complements of the risky option. The results are more consistent with fuzzy-trace theory than with prospect theory. The consequences of these findings for the application of formal models like prospect theory, and for rationality, are discussed.