This article reviews the strengths and weaknesses of “surrogate selection” as a solution to intractable medical futility disputes. It concludes that while surrogate selection is an increasingly viable solution, it remains only a partial solution because it is often difficult or impossible to demonstrate that a surrogate demanding non-recommended end-of-life medical treatment is acting outside the scope of her authority. Over the past twelve years, many states have been developing new legislative solutions to intractable medical futility disputes. The most widely-discussed solution empowers healthcare providers to unilaterally refuse patient- or surrogate-requested treatment that the provider deems inappropriate. In Texas, for example, when providers and patients’ surrogates cannot agree on appropriate life-sustaining medical treatment, the state Advance Directives Act designates the hospital ethics committee as adjudicator of last resort. The Idaho Senate recently passed a bill to the same effect. And policymakers in other states are considering similar legislation. But Harvard Medical School Professor Robert Truog argues that this too-provider friendly sort of internal dispute resolution legislation is h unnecessary and dangerous. He argues that current healthcare decisions laws in every state already give healthcare providers a mechanism to avoid providing inappropriate medicine demanded by surrogates. Specifically, when a surrogate demands treatment that providers deem medically inappropriate, these “surrogate selection” laws often permit providers to designate a new surrogate from whom consent to withhold/withdraw can be obtained. The logic behind these laws is simple. A surrogate is the patient’s agent and, as such, must act according to the patient’s instructions, known preferences, and best interests. When a surrogate exceeds the scope of her authority, she can and should be replaced. For example, providers took a surrogate selection approach in the famous Helga Wanglie case. Rather than ask the court to make the treatment decision, providers asked the judge only to make a procedural decision appointing a new decision maker. The court-appointed guardian, in turn, would make the substantive decision. While providers were ultimately unsuccessful in both Wanglie and similar cases during the early 1990s, surrogate selection has, more recently, proven to be a successful approach. Emblematic of this approach, in a string of published cases during 2008 and 2009, courts from New York to California replaced patients’ surrogates because they were demanding inappropriately aggressive end-of-life treatment. Yet, while surrogate selection has become increasingly well-grounded in both statutory standards and judicial precedent, it remains only a partial solution to intractable futility disputes. To replace a surrogate requires evidence of a contradiction between the surrogate’s decision and the patient’s instructions, known preferences, and best interests. But, for two main reasons, providers will frequently be unable to demonstrate any such contradiction. First, providers will often lack enough evidence of patient instructions or preferences to demonstrate surrogate deviation. Second, in many cases, surrogate decisions (for non-beneficial treatment) will actually be in harmony with patient wishes. In sum, Professor Truog is right to endorse surrogate selection as a solution to intractable futility disputes. It is a mechanism that can often work, and one that, where possible, should be preferred over power-shifting laws. Yet, since surrogate selection cannot resolve significant categories of conflict, we must still develop dispute resolution mechanisms to handle those remaining disputes in which providers conflict with “irreplaceable” surrogates.