Many believe that the treatment‐enhancement distinction (TED) marks an important ethical boundary that we should use to shape public policy on biomedical interventions. A common justification for this purported normative force appeals to the idea that, whereas treatments respond to genuine medical needs, (most) enhancements can only satisfy mere preferences or “expensive tastes”. This article offers a critique of that justification, while still accepting the TED as a conceptual tool, as well as some of the key ethical axioms endorsed by its proponents. I begin by laying out the TED, the practical implications that tend to be drawn from it, and the justification just sketched for these implications. Using examples drawn from preventive medicine, biomedical technology, and other categories of biomedical interventions, I then go on to challenge both the presupposition of a fundamental dichotomy between treatments and enhancements, and the assumption that enhancements – barring rare exceptions – cannot serve legitimate medical needs. Finally, I consider some ways in which supporters of the TED might try re‐formulating the distinction to blunt the force of my critique. I conclude that such a move cannot fully succeed, and that while the TED does have some degree of normative force, it nevertheless cannot play the role that its advocates expect from it. Seeking to justify a general presumption against enhancements based on the rationale I critique here would mean ignoring their various potential benefits, including medical or therapeutic, and would reflect a prejudice – which I refer to as “treatment fetishism”.