The extent to which teachers and school directors are held to account may play a central role in determining education outcomes, particularly in developing and transition countries where institutional deficiencies can distort incentives. This paper investigates the relationship between an expanded set of school inputs, including proxies for the functionality of "top-down" and "bottom-up" accountability systems, and education outputs in Albanian primary schools. The authors use data generated by an original survey of 180 nationally representative schools. The analysis shows a strong negative correlation between measures of "top-down" accountability and students' rates of grade repetition and failure in final examinations, and a strong positive correlation between measures of top-down accountability and students' excellence in math. "Bottom-up" accountability measures are correlated to various education outputs, although they tend lose statistical significance once parent characteristics, school resources and top-down accountability indicators are considered. An in-depth analysis of participatory accountability within the schools focuses on parents' willingness to hold teachers to account. Here, the survey data are combined with data from lab-type experiments conducted with parents and teachers in the schools. In general, the survey data highlight problems of limited parental involvement and lack of information about participatory accountability structures. The experiments indicate that the lack of parental participation in the school accountability system is owing to information constraints and weak institutions that allow parent class representatives to be appointed by teachers rather than elected by parents. Appendices include: (1) Design and Sampling Strategy for the Albania School Stakeholder Survey and Field Experiment; (2) Statistics from the Albania School Stakeholder Survey, by RED; and (3) Design and Application of the Field Lab-Type Experiments (LTEs). (Contains 4 figures, 17 tables and 50 footnotes.)