This dissertation develops a model investigating how computerization shapes the distribution of economic activities across space. Computer capital is a complement to complex tasks and these tasks generate and benefit from knowledge spillovers in large agglomerations. Thus, the complementarity between computer capital and city size is skill-biased where workers with a comparative advantage in those tasks disproportionately sort in large cities. Data from the United Kingdom and Germany support a complementarity between computer capital and city size. Indeed, an increase in city size is associated with a significant increase in the probability of different indicators of computer capital: using a computer at work, having computerized equipment as the most important work equipment, having an advanced use computer at work. The positive relationship holds even controlling for detailed workers and jobs characteristics. Consistent with housings costs being the dispersive force linked to computerization for low-skilled workers, the complementarity between computer use at work and city size is positively significant for them also. Finally, I develop a structural model where parameters are estimated empirically to quantify the impact of computerization on the increasing sorting of high-skilled workers in large cities.