This thesis considers three ways research on cognition can be used to improve our understanding of development economics. The first chapter explores whether cognitive science research on the development of mathematical abilities can be used to design interventions to improve learning in resource-poor contexts. We developed a series of games designed to train children’s intuitive concepts of number and geometry and evaluated their effectiveness with a randomized field experiment in Delhi, India. We found that the intervention produced enduring improvements in the exercised abilities, but that these improvements did not translate into improved formal mathematics skills. The second chapter asks whether the omnipresent noise in developing contexts reduces worker productivity by impairing cognitive function. Using a pair of randomized field experiments, I found that noise substantially reduces productivity, that this appears to occur through the proposed cognitive mechanisms, and that workers do not seem aware of these effects. The final chapter evaluates whether there is scope for cognitive biases to affect the adoption of energy efficient technologies in developing contexts.