In April 2007, more than 60 people--practitioners, advocates, funders, and evaluators--met for 2 days in Pittsburgh, Pennsylvania, to discuss the challenges that arise in evaluating systems change or systems-building initiatives and to suggest approaches to resolving those challenges. The symposium was convened by the partners of BUILD, a multistate initiative aimed at ensuring that children from birth to age five are safe, healthy, eager to learn, and ready to succeed in school. The Pittsburgh symposium was organized for two reasons: (1) to help the BUILD evaluators identify strategies for addressing the specific problems they are facing; and (2) to take advantage of the intensive focus on BUILD to illuminate evaluation issues that the stakeholders are likely to face in any ambitious initiative that aims to improve public systems. This article describes the BUILD symposium generally and identifies several of the important themes that emerged in the 2-day discussion. The most important theme revolved around problems identifying the theory of change and the theory of action for systems building. The evaluation-specific issues related to limiting and targeting the scope of the evaluation, getting clarity about what outcomes can and cannot be expected at which points in the initiative's evaluation, and ensuring that the evaluation serves learning objectives for the practitioners and for the field. (Contains 1 note.)