<p>Accessibility to public facilities and urban services is a crucial determinant of neighborhood sustainability and a central aspect of spatial justice in cities. However, conventional accessibility assessment methods often rely on rigid data classification, leading to a loss of geographic information and analytical accuracy. This study aims to address this gap by integrating Geographic Information Systems (GIS) with a Fuzzy Inference System (FIS) to provide a more flexible and precise framework for accessibility evaluation. The proposed methodology was applied in Urmia, Iran, using detailed spatial and demographic data on parks, educational, health, religious, and sports services. This study contributes to accessibility research by developing a transferable hierarchical GIS–FIS framework that (i) models neighborhood-level accessibility to <i>multiple</i> urban services simultaneously, (ii) preserves continuous spatial information by using network-based pedestrian distances and kernel-density population inputs, and (iii) aggregates service-specific fuzzy outputs through a higher-level FIS, enabling robust identification of underserved neighborhoods for urban policy interventions. The FIS model incorporated both distance and population density as key inputs, producing neighborhood-level accessibility indices. Results show that neighborhoods 32, 49, and 50 have the lowest accessibility to urban services, while neighborhoods 21, 23, and 26 demonstrate the highest levels. Comparative analysis highlights that the integrated GIS–FIS approach reduces uncertainty inherent in conventional methods, avoids information loss due to data classification, and enables more realistic representation of spatial inequalities. These findings emphasize the potential of fuzzy logic–based accessibility modeling as a robust decision-support tool for promoting spatial justice in rapidly growing cities.</p>