<![CDATA[An alternative monitoring tool of fine particulate matter (PM₂.₅) is required to support air quality management. It is due to the limited availability and high cost of high-standard instruments that encouraged the use of low-cost sensors based on the Internet of Things. This study aims to evaluate the accuracy of the PMS5003 and PMS9003 sensors against reference instrument of the Met One ES-642 operated by the Jambi City Environmental Agency, as well as to analyze the relationship between temperature and relative humidity with the measurement results. The study used a quantitative design with field observations conducted over 25 days during April–May 2026. Evaluation was carried out using descriptive statistics, Spearman correlation test, Friedman test, coefficient of determination, MBE, MAE, RMSE, and conversion of PM₂.₅ concentration into the Air Pollutant Standard Index (ISPU). The results show that both sensors were able to capture the pattern of PM₂.₅ concentration changes, but tended to produce higher values than the reference instrument. PMS5003 had an MBE of 6.17 µg/m³, MAE of 17.34 µg/m³, RMSE of 23.53 µg/m³, and R² of 0.12, while PMS9003 had an MBE of 10.12 µg/m³, MAE of 18.40 µg/m³, RMSE of 27.16 µg/m³, and R² of 0.11. Relative humidity showed a moderate positive relationship, while temperature showed a weak negative relationship. Low-cost sensors have the potential to be used for indicative monitoring, but their interpretation must take measurement bias and humidity conditions into account.]]>
