AbstractSince March 2020, the COVID-19 disease has become a global concern, and its concentration has been primarily in urban settings. Previous research suggests that multidimensional factors allow understanding the distribution of the disease but has limitations such as having nonhomogeneous units as the object of study, not incorporating changes in sanitary control measures over time or the absence of mobility variables. To overcome these shortcomings, we investigated the association between socioeconomic, demographic, and built environment factors with infection rates in the Metropolitan Area of Barcelona, one of the most compact and mixed-use environments in Europe. For this purpose, we use spatial regression models at five different stages that capture variations in sanitary control measures. Our results indicate that before the lockdown, infections were concentrated in high-income areas, but once it started the pattern shifted toward areas characterized by overcrowding, with more people who did not have the opportunity to telework, as well as nursing homes. Although commuting time also maintained a positive association with infections, the use of public transportation was not observed to have a direct impact. Contrary to what was speculated at the beginning of the pandemic, density was not shown to be a decisive factor in explaining infection rates; therefore, the results suggest keeping the focus on the quality of housing to avoid intrafamily infections but particularly in those where elderly dependents live. Likewise, public transportation can maintain its benefits for the most vulnerable urban populations as long as minimum safety measures are guaranteed in its interior.