A quantitative assessment of preschools in São Paulo (Brazil), with a focus on walking and the right to the city. Abstract (793 words) In the last decades, much has been written about the role of urban mobility to achieve social justice. However, changes in public policies towards territorial equity have found increased opposition not only in the public sector (from municipalities to the national government), but also in civil society (citizens, corporations and social movements), with important decision makers increasingly overlooking the impacts of transport on climate change, public health and socio-spatial disaggregation, among others. In this sense, public agencies and advocacy-led organizations have identified in schools a set of relevant factors to increase the right to the city through the promotion of permanent changes in transport practices. These refer mainly to the initiatives regarding the promotion of active mobility, which includes walking to relevant destinations, such as school. The study here presented seeks to examine and conceptualize the vectors of mobility capabilities and functionings of preschools and nurseries (children aged up to 6 years old) in São Paulo, Brazil. Despite being identified as an appropriate tool to address different dimensions of transport-related social exclusion and inequalities, the Capability Approach lacks application and conceptualization within the urban mobility domain, especially regarding the case of children and the access to school in Latin American cities. To pursue the identification of differences in the accessibility levels to schools through a quantitative approach, a georeferenced dataset containing public and private preschools and nurseries in the municipality of São Paulo (sample of 203 schools, of which 109 are public) underlay the generation of circular buffers (Euclidean distance) and service areas (along the street network) around schools with diameters of 500m, 1,000m and 2,000m, which correspond to 10, 20 and 40-minute walking distances, respectively, once a preschool-age child is considered. From the location of schools, information from georeferenced datasets was incorporated relating to i) mobility, including the share of pedestrian trips and the average walking time to school; ii) the street network, such as the connectivity of the street network in the school's surrounding area (street connectivity), and the centrality of the school's location in the urban street network (betweenness centrality); iii) the road safety, i.e. the rate of pedestrian-car crashes during school drop-off and pickup ; and iv) the quality of the built environment in the surroundings of schools, which refers to the population density, the social vulnerability, the concentration of preschool-age children, income distribution, among others. To integrate capability metrics into a composite measure that indicates the capability of schools to promote walking mobility, we identified the relevant thresholds for each metric to classify the schools as capable or non-capable of accessing preschool and nurseries through walking. In this way, the obtained indicator of mobility capabilities enabled the classification of schools in five levels of mobility capabilities: i) Extreme difficulty; ii) Severe difficulty; iii) Moderate difficulty; iv) Mild difficulty; and v) No difficulty. Each level of mobility capability was associated with a defined range of values of the indicator of mobility capabilities. Once grouped into five levels of mobility capabilities, we could evaluate the mobility agency of each school, i.e., the corresponding capacity to promote pedestrian mobility (functionings). For that, the expected shares of pedestrian trips to school (upper and lower bounds) were assigned to one of the five levels of mobility capabilities. Accordingly, the higher the level of mobility capability of a school (i.e., the better the street network, traffic safety and the built environment for walking), the more capable was the school's environment found to promote higher shares of pedestrian trips to and from schools. So, a school's mobility agency could be obtained by subtracting the current share of pedestrian trips from the closest value of the expected share of pedestrian trips, which could be either positive (above the upper bound) or negative (below the lower bound). Results pointed out that public schools present share of pedestrian trips two times higher and average walking time 30% higher than private schools. Schools with positive agency (mostly the public ones) requires investments on built environment of the surroundings of schools towards providing better conditions for walking trips of the children and their caregivers. On the other hand, schools with negative agency (mostly the private ones, which are generally located in places with less barriers to walk) might require educational programs for motivating higher levels of use of active modes. Besides the contribution in the relationship between the potential and the effective practices of walking to school from the Capability Approach, the study here presented aimed at providing insights on the accessibility gaps in public and private preschools and nurseries, especially the ones related to the community severance and the physical constraints of the built environment. In this sense, the study can be helpful when drafting public policies to promote walking as a primary way to access public and private schools.