This chapter focuses on the substantial gaps in school enrollments and cognitive achievement between boys and girls in rural Pakistan and between a relatively prosperous district (Faisalabad, Punjab) and a poor district (Badin, Sind). An important policy question is whether the allocation of educational expenditures to target disadvantaged groups will assure more equitable schooling outcomes. Surveys conducted in rural Pakistan provided data on literacy, numeracy, educational achievement, preschool reasoning ability, parental schooling and income, student school costs, and school quality. Data on 637 respondents aged 10-25 were used to assess the relative impacts of targeted improvements in schooling quality versus school quantity. By using estimated earnings functions that translate cognitive skills gains into income gains and by linking income gains to estimated costs of improvements, rates of return on investment were generated for: (1) providing a low-quality primary school (K-5) where none was available previously, (2) raising primary school quality from low to high, and (3) enabling the graduate of a low-quality primary school to complete middle school. Findings indicate that local school availability accounted for 40 percent or more of gender and regional gaps in cognitive achievement. Making even low-quality primary schools available had the highest return for all groups except girls. In all cases, raising the quality of primary schools had a substantially higher rate of return than expanding middle-school availability. Targeting educationally disadvantaged groups would involve little, if any, sacrifice in aggregate returns to society in Pakistan. Contains 55 references. (SV)