President Obama has proposed "Preschool for All" with a pledge of $75 million to the states to implement "high quality" full-time preschool programs for low-income children as part of the regular K-12 system. The states would come up with additional funds to provide for the remaining higher income students. This proposal comes despite the fact that rigorous evaluations of the existing federal pre-K program, Head Start, show no lasting effects. Presumably, Head Start does not have sufficient quality, at least as compared to several "high quality" programs implemented in various states and school districts such as New Jersey, Oklahoma, Boston, and Tennessee. This paper reviews existing research on Head Start and these selected "high quality" programs. It finds that the quality of Head Start programs is actually higher than the New Jersey program, and moreover there is little evidence that common measures of quality are strongly related to preschool outcomes. This makes sense when one considers that pre-K children are only 3 or 4 years old, and the necessary subject matters can be mastered by instructors without formal college degrees. The Tennessee program used a randomized design similar to Head Start, and it did not find significant long-term outcomes. Other evaluations of high-quality pre-K programs use a research design (knows as RDD) that is susceptible to serious bias due to program dropouts. The paper recommends that a small portion of Head Start funds be used to develop a thorough and rigorous demonstration and evaluation of "high quality" preschool programs before spending billions of dollars on another untested educational program. (author abstract)