The study examined the concentration, size distribution, and seasonal variations of EPFRs in the North China Plain region, as well as investigated their sources using PMF. It also explored the role of EPFR speciation in contributing to the oxidative potential of PM. I find this study very interesting and important. The authors have done a great job to discuss other studies comprehensively. I only have two comments as follows:

1. The conclusion that the majority of EPFRs are present in PM2.5 is drawn from the fact that the average EPFRv in PM2.5 accounts for over 95.2% of those in PM10 and TSP. However, the box plots in Fig. 1 suggest that PM2.5 EPFRs may not make up such a high fraction. I recommend that the authors calculate the fraction of EPFRs in PM2.5 in PM10 or TSP for each sample and average it for discussion, for better representativeness.
2. In Section 3.3.2 and Fig. 4, the associations of OP with various chemical species are discussed. The OP values are in mass-normalized activities, while the chemical species are in ambient concentrations in m3. It would be helpful for the authors to explain why volume-normalized OP data were not utilized for the association discussion, which makes more sense to me.

This study presents measurements of environmentally persistent free radicals (EPFRs) and oxidative potential (OP) of different size fractions of PM (PM_2.5, PM₁₀, TSP) collected over 1 year in a rural site in the North China Plane. The authors investigated the sources of EPFR using positive matrix factorisation and explored the role of solubility on EPFRs and their contribution to OP. Overall, the study is well written and presented, providing new insights into the sources contributing to EPFR, as well as the influence of EPFRs on OP, and the results are comprehensively compared to the literature. The findings in this study are novel and certainly within the scope of ACP. I recommend publication and only have a few minor comments.

• Line 57 - I am unclear as to the meaning of nonsolvent-extractable, please clarify.
• Were filters in this study all extracted and analysed at the end of the sampling campaign or were they systematically analysed during the measurement campaign? This is important to clarify given the broad range of lifetimes of EPFRs.
• References for the DTT protocol should be provided.
• Line 121 – Typo “The rest DTT...”
• Line 122-126. The authors state that 2-OHTA production is proportional to OH formation, however Gonzalez et al (2018) demonstrated that the yield of the 2-OHTA from the terephthalate-OH reaction is around 31%, and thus calculated OH concentrations need to be corrected concerning the yield of 2-OHTA . Have the author’s considered this when calculation OP_OH? Ref: https://doi.org/10.1080/00032719.2018.1431246
• Regarding the correlation of OP to components (Section 3.3.2, Figure 4), mass normalised OP values were used, but were these correlated with mass normalised components (e.g. Fe ug^-1)? This should be clarified. It also seems strange that correlation analysis between EPFR and components is not presented in the main manuscript in Figure 4, given a lot of the manuscript focus is on EPFRs.