The existing experimental evidence on returns to capital is based on randomisation of either interest-bearing loans or grants. This paper reports on a field experiment conducted in Pakistan in which interest-free loans were randomly provided to microenterprises. The study was conducted in co-operation with Akhuwat Islamic Microfinance. We find that treatment leads to a substantial increase in working capital and in business profits. Using randomised treatment as an instrument for capital, we find average monthly returns to capital of 8.6 to 11.9 percent a month. These returns are higher than the interest rates charged by conventional microfinance institutions in Pakistan.