Peatland fires in Indonesia drive severe air pollution. Studies have focused on El Niño years, therefore neutral and La Niña years remain uncharacterised. We deploy a dense network of PM_2.5 sensors across Central Kalimantan peatlands between August 2023 and October 2025 to quantify how El Niño–Southern Oscillation (ENSO) modulates the magnitude and spatial variability in dry season PM_2.5 concentrations. Sensors were installed at urban, rural, and remote locations, spanning El Niño (2023), neutral (2024), and La Niña (2025) dry seasons. During the 2023 El Niño dry season, low rainfall and deep water tables enhanced peatland flammability and supported extensive peat fires. Urban and rural sites exceeded the WHO 24‑hour PM_2.5 guideline on 99 % and 97 % of days. A remote site exceeded these guidelines on 85 % and 24 % of days, with fire smoke influence confirmed by low spatial variability and a dual‑peak diurnal cycle across sites, indicating regional pollution rather than local sources. As ENSO conditions shifted to neutral (2024) and La Niña (2025), increased rainfall and shallower water tables reduced fire activity and PM_2.5 concentrations. In 2024, WHO guideline exceedances fell to 41 % and 12 % at urban and rural sites. Our results indicate that even across non-El Niño years, dry season PM_2.5 is influenced by regional fire emissions, but the magnitude and spatial variability is modulated by ENSO-phase. Our results demonstrate that dense sensor networks can distinguish regional fire smoke from local pollution, enabling early detection of fire‑driven air quality degradation. Reducing peatland fires through restoration and fire management would deliver consistent air quality benefits.