Within America's school systems, sometime between kindergarten and secondary education, a wide variation appears among the achievement levels of different children. The learning gap between high-achieving high schoolers and dropouts is certainly no secret to educators. Huge sums of federal funds and foundation support have been injected into K-12 education in an attempt to bridge the learning gap. From the Annenberg Foundation's millions to the billions attached to Title I, from whole-school restructuring efforts to class-size reductions, from two decades of effective schools research to the Blue Ribbon Schools, Goals 2000 and generous Gates Foundation money, no simple solution has surfaced for bridging this gap and leaving no child behind. The author argues that the answer to bridging the learning gap lies not in asking the question "How do children learn?" but reframing it as "When do children learn?" Technological advancements and the relatively new field of cognitive neuroscience are helping to shine a bright light on the learning problem. Positron emission tomography scans and functional magnetic resonance imaging are helping Americans to answer this more critical and promising question of when children learn, and the implications are startling. The earliest years of life, before a free and appropriate education is available, appear to hold the key to learning. The author stresses a truly seamless free and appropriate education for all children should begin on the maternity ward. The author contends that only through redefining America's educational boundaries can Americans expect to close the learning gap.