Since 1980, the number of twin births in the United States has increased 76 percent, and the number of triplets or higher‐order multiples has increased over 400 percent. These increases are due in part to increased maternal age, which is associated with spontaneous twinning. But the primary reason for these increases is that more and more people are undergoing fertility treatment. Despite an emerging (but not absolute) consensus in the medical literature that multiples, including twins, should be a far less frequent outcome of fertility treatment, American clinicians currently practice fertility medicine in ways likely to result in twins and, occasionally, triplets—often, they tell us, at the request of their patients. At first blush, one might conclude that these practice patterns show that patients and their doctors have weighed the risks associated with twins against the benefits of swiftly completing their families and decided that the risks are worth taking. That is, their requests for twins must represent their free and informed choices. But there are reasons to be skeptical of this interpretation, including that the preferences of many patients seem to be very strongly shaped, if not sometimes completely constrained, by concerns about costs—costs that mainly result from the routine exclusion of in vitro fertilization from health insurance coverage. It is quite possible, we believe, that these cost concerns actually undermine the autonomy of fertility patients, pushing many of them to take risks with their health and the health of their hoped‐for future children.