Despite the tremendous expansion in education access worldwide, countries differ dramatically both in primary and secondary enrollment rates and in student achievement. Although per capita income explains a great deal of the difference, schooling outcomes vary sharply even among countries at similar income levels. This study asks whether developing countries with relatively better and worse secondary schooling enrollment and achievement outcomes differ systematically in the structure of their education finance (education expenditure as a share of gross domestic product (GDP), secondary education expenditure as a share of total education spending, and per student costs) and service delivery features such as pupils per teacher, percent of students repeating a grade (the repetition rate), and percent of students older than expected for their grade level (the overage rate). It may well be that countries with limited secondary education have dedicated scarce resources to other priorities, such as the expansion of primary education or investments in infrastructure, and the author does not wish to imply that these countries should necessarily divert scarce resources to the expansion of secondary education. At the same time, a common finding in the literature is that unit costs appear to be lower in school systems with better outcomes. As such, it may be the case that countries can improve secondary education outcomes without a large infusion of resources. Using available data, the author argues that it is worthwhile to examine how countries with better outcomes at similar income levels manage to perform as well as they do. Several studies have explored these questions at the primary level and at least one has addressed the secondary level. In addition to the question of whether more successful systems bear lower unit costs, the literature also considers differences in pupil-to-teacher ratios and repetition rates. The author's study contributes two innovations. First, she supplemented the standard outcome measure of net enrollment rates (NERs) with student test scores from the 1999 Trends in International Math and Science Study (TIMSS). Second, she identified better-performing countries as those with better-than-predicted outcomes based on their income. By examining the extremely close association between income and schooling outcomes and the considerable variation in outcomes among countries at the same income level, the author was able to identify policies that do not owe their success to being implemented in higher income environments. (Contains 1 figure, 6 tables and 14 footnotes.)