Walkability is a factor that affects property valuation because a higher pedestrian flow leads to a higher valuation of commercial properties. Meanwhile, residents might prefer to live on quiet blocks. Balancing these aims has been essential since existing major cities were built in early modern times. This paper studies the return on housing rent in the City of Osaka, Japan, in early modern times, focusing on bridges that were corners of pedestrian flows and raised walkability of neighboring blocks across channels for transportation. By means of archival work and regression analysis on the firsthand documents of the largest landlord, Mitsui, we find first that the more distant from a bridge a property was, the higher its rent was, and second that the closer to a bridge a property was, the higher its profit rate was, controlling for property characteristics, the city’s population, consumer prices, interest rates, and gross domestic product. Properties in inner quiet blocks were more valued in terms of rents, but properties close to corners of pedestrian flows delivered higher profits. R-Learner, a machine learning semiparametric algorithm, supports the results.