Improvement in agricultural productivity plays a key role in the process of economic development. Investment in critical infrastructure has been documented in the literature as one of the pathways to boost agricultural productivity. In this paper, we ask whether foreign aid aimed at economic and social infrastructure can help improve agriculture productivity in Tanzania. To do so we combine household panel data with rich farm level information with geocoded Chinese development projects. Since Chinese projects are not randomly allocated among villages, we first balance the treated and control villages using the Entropy balancing method based on a set of village level characteristics. We then compared the productivity of farms located in villages closer to at least one Chinese project prior to the year of survey to those located farther away in a diff-in-diff framework. Preliminary results indicate that agricultural land and labor productivity are higher in villages close to an active Chinese project.