There is debate about measurement of impact and achievement of expected results of rural microfinance programs in LDCs. The paper addresses these issues using survey household data from Bolivia. Regression models examine the determinants of education outcomes for microfinance clients. The results challenge usual assumptions in program design.