["Geographical Analysis, Volume 58, Issue 2, April 2026. ", "\nABSTRACT\nTraditional measures of urban accessibility often rely on static models or survey data. However, location information from mobile networks enables large‐scale, dynamic analyses of how people navigate cities. In this study, we employ eXtended Detail Records (XDRs) from mobile phone activity to analyze commuting patterns and accessibility inequalities in Santiago, Chile. We identify residential and work locations and model commuting routes by public transport and walking using the R5 multimodal routing engine. Spatial patterns are examined using bivariate local indicators of spatial association (LISA) alongside regression techniques to identify distinct commuting behaviors and their alignment with vulnerable population groups. Our results show that while average public transport commuting times do not differ significantly across socioeconomic groups, marked inequalities emerge when accessibility is considered. High‐income neighborhoods consistently exhibit high accessibility, whereas low‐income areas show substantially lower levels. Importantly, these disparities do not translate into longer commuting times for lower‐income groups, indicating a weak relationship between proximity to opportunities and observed travel times. The analysis also reveals significant disparities across sociodemographic groups, particularly in relation to Indigenous populations and gender. The proposed approach is readily scalable and can support evaluations of changes in commuting patterns and the impacts of urban interventions.\n"]