Spatial segregation is a pressing issue in Latin American cities due to high urbanization, population density, and inequalities. This study examines segregation patterns in 84 middle to large cities in Colombia and their relationship with the spatial configuration of cities using satellite data. The analysis focuses on evenness and isolation metrics of segregation for ethnicity, education level, and employment status. The findings reveal higher segregation in education level and race/ethnicity in terms of evenness dimension, and in education level and employment status in terms of isolation dimension. The study also identifies associations between segregation and built environment metrics. Negative associations are observed between education level and race/ethnicity segregation and fragmentation, as well as between employment status and patches of isolation. Positive associations exist between education level and fragmentation, race/ethnicity and patches of isolation, and a negative association between employment status and street density. These results have implications for public urban policies, particularly in small and medium cities, where there is limited understanding of segregation dynamics. The study highlights the need to consider factors beyond income, such as ethnicity, in addressing spatial segregation in urban planning and policy-making.