Abstract: Hourly continuous observations of the chemical compositions of fine particles and gaseous precursors were conducted during autumn and winter haze pollution episodes in urban Beijing. The ISORROPIA II model was used to simulate particle pH, and the driving factors of particle pH were studied using a sensitivity analysis method. The results show that the average particle pH calculated using the ISORROPIA II model is 4.52±0.52, 5.19±0.28, and 5.03±0.79 for three heavy haze pollution events in the autumn and winter seasons of 2017, 2018, and 2019, respectively. From the cleaning stage to the moderate pollution stage and from the moderate pollution stage to the severe pollution stage, the average value of particle pH decreases by 0.72–1.07 and 0.3–1.03, respectively, showing a gradually decreasing trend with the aggravation of haze pollution. The sensitivity tests demonstrate that particle pH exhibits different responses to the changes in chemical composition. Specifically, the particle pH is mainly affected by TNH_x total ammonium (gas+aerosol) and SO₄²⁻, with limited impact from changes in ambient temperature and humidity. Particle pH is more sensitive to TNH_x than SO₄²⁻ and NO₃⁻. In addition, high TNH_x levels reduce the sensitivity of particle pH to SO₄²⁻ (TNH_x＞47 μg/m³) and NO₃⁻ (TNH_x＞30–40 μg/m³). During heavy haze pollution in Beijing, the atmosphere is rich in ammonia; therefore, particle pH increases to ＞4. Furthermore, the results indicate that particle pH and the magnitude of the change in particle pH are related to the TNH_x level and the sensitivity of particle pH to SO₄²⁻ and NO₃⁻ in heavy haze pollution events. These relations can serve as an effective reference for clarifying the formation mechanism of secondary aerosols.