Accurate measurement of returns to education is particularly important in African countries where government resources are extremely limited, and education's competition for budget dollars is intense. This paper attempts to measure consistently the returns to education in Ghana. The chief problem to be handled in the measurement of returns to education in Mincer-based wage regressions is ability bias. This paper attempts to address this problem from a completely new perspective--that of the firm. It begins by specifying a production function that is consistent with the Mincerian wage equation. The production function is estimated consistently, taking into account the simultaneity of labor demand and output decisions. A result of including worker schooling, experience and ability in the production function specification is that a measure of worker ability, defined as the worker's contribution to firm product after controlling for schooling and experience, is acquired. This measure of worker ability can be used in a wage equation in order to control directly for worker ability to obtain consistent estimates of the returns to schooling. The model is estimated using data from the manufacturing sector in Ghana, but this technique is replicable to other countries and datasets, given the recent increase in available linked employer-employee datasets.