This paper analyses the eects of bank lending on GDP and employment. Following losses on international nancial markets in 2008/09, a large German bank cut its lending to the German economy. I exploit variation in dependence on this bank across counties. To address the correlation between county GDP growth and dependence on this bank, I use the distance to the closest of three temporary, historic bank head oces as instrumental variable. The results show that the eects of the lending cut were persistent, and resembled the growth patterns of developed economies during and after the Great Recession. For two years, the lending cut reduced GDP growth. Thereafter, aected counties remained on a lower, parallel trend. The rm results exhibit similar dynamics, and show that the lending cut primarily aected capital expenditures.Overall, the lending cut reduced aggregate German GDP in 2012 by 3.9 percent, and employment by 2.3 percent. This shows that a single bank can persistently shape macroeconomic growth.