I estimate an elasticity of wages with respect to city population of about 5 percent for Colombian cities. This finding is robust to a number of econometric concerns. The second main finding is a negative effect of market access on wages. Third main finding regards stronger agglomeration effects in the informal sector. In turn, this explains a range of other negative findings, including only weak evidence in favor of human capital externalities, no evidence of a complementarity between cities and skills, and an absence of learning effects. I do not find measurable effects of roads or amenities on wages either.