We study whether information provision improves the academic performance of primary school children in a setting where parents have incomplete information about their child's cognitive skills and there exist competing public and private providers of education. Contiguous village councils in the north Indian state of Rajasthan were randomly assigned to either a control or one of four treatment groups in which schools and/or parents were provided information through report cards on either intra or both intra and inter school performance of students in curriculum based tests. We find significant improvement in test scores of private school students by 0.31 standard deviations when information on both absolute and relative school quality is provided to households and schools. There are no significant improvements in the learning outcomes of public school children in any treatment. Close examination of the results suggest that private school students chose better quality schools in the new academic year. Public school parents did respond by exercising school choice and lowering student absenteeism but saw no improvements in learning outcomes possibly because of constrained school choice set. Overall, our results suggest that information on the relative quality of schools can be a cheap and effective tool for improving learning outcomes when households can exercise school choice.