"Education for All" (EFA) identifies goals and targets and translates these into indicators which are used to evaluate progress and influence flows of external financing. The search for evidence based policy depends on measures of performance that can link cause and effect and that represent real gains in progress towards desired outcomes. However, the indicators widely used for access to education have serious problems. Gross and Net Enrolment Rates (GERs and NERs) are widely used to assess participation, but their changing values can provide misleading signals to policy makers. Grade specific enrolment and completion rates may be better but are silent on issues of quality and achievement. Gender Parity Indices (GPIs) foreground differences related to children's sex but often aggregate and conceal underlying patterns of participation (by age, location, household income) and may make invisible seriously unequal numbers of boys and girls in populations. The Education for All Development Index (EDI) suffers from being a highly aggregated composite index which is difficult to interpret and of limited use since changes over time are often within its margins of error. This paper selects some of the key indicators used for EFA and offers an analysis of their strengths and weaknesses. Improved target setting with better indicators is needed which gives a more nuanced understanding of what indicators do and do not measure. (Contains 5 tables and 5 figures.)