<![CDATA[Purpose of the study: Access to safe, accessible, and disability-responsive physical education (PE) environments is part of educational inclusion, yet school-level evidence on PE facilities for students with disabilities remains limited in many low- and middle-income settings. This study mapped PE and health-related facilities across all Special Schools (SLB) operating in Pekanbaru, Indonesia, and identified priority deficits. Materials and methods: A descriptive cross-sectional census survey was conducted in all five officially operating SLBs in Pekanbaru during May–June 2026. Seventeen PE teachers assisted as key informants. Data were collected through direct observation, photographic documentation, and a 12-indicator structured checklist covering spaces, equipment, accessibility, safety, and sanitation. Availability was calculated as observed provision relative to the target standard, capped at 100%. School-level scores were summarized using means and standard deviations; a descriptive 95% confidence interval was calculated for the city-wide mean. No inferential hypothesis testing was performed because the five eligible schools constituted the complete accessible population. Results: Mean availability was 78.81% (SD 14.19; descriptive 95% CI 61.19–96.43), classified as Complete under the study's five-category scale. Scores ranged from 66.67% to 100.00%. Motor-coordination equipment, adaptive game equipment, access paths/wheelchair ramps, and disability-friendly toilets reached 100% across schools. The lowest indicator scores were mats (37.14%), whistles/stopwatches (40.00%), and sports fields (60.00%). Conclusions: The census shows that aggregate completeness can coexist with substantial deficits in safety-critical and instructional resources. Internationally, the study contributes context-specific evidence to the ecological view that school physical environments form part of the opportunity structure for physical activity and inclusive PE. The findings support targeted infrastructure planning, while the small census frame and non-validated checklist limit generalization and causal interpretation]]>
