This study employs the iodide-adduct chemical ionization mass spectrometer to achieve qualitative and quantitative measurements of bicyclic peroxy radicals, key intermediates in aromatic oxidation. Three vital instrumental parameters influencing the sensitivity to BPRs were systematically identified and optimized. Then it uses direct BPRs quantification to assess uncertainties in the traditional product-yield method for elucidating aromatic oxidation mechanisms. The calculation process is rigorous, and the quantitative data quality is high, making the work a valuable contribution to understanding atmospheric oxidation processes. The manuscript is recommended for acceptance after minor revisions. Specific comments are as follows:

• Line 100: The authors stated that compounds with similar m/z values and functional groups (e.g., hydroxyl group) were selected to serve as proxies for BPRs. More error analysis would be better here.
• Line 116: Why was a residence time of 3~4 seconds chosen in the experiment? Please elaborate on the impact of the residence time on the calibration process.
• In section 2.2, it is mentioned that I^- is similar to Br^- and can be used to measure HO₂. Has it been considered whether using HO₂ radicals instead of BPRs is feasible in terms of reactivity, or whether possible error estimations can be made?
• Although figure space is limited, it is necessary to label each bar in Figure S1 with the corresponding species or to provide an extra accompanying table.
• The calibration section noted that low NO levels were chosen to avoid excessively complex RO₂ However, aromatic compounds are predominantly anthropogenic and are usually accompanied by elevated NOx concentrations in the real atmosphere. Please add a statement addressing this limitation in the manuscript.

General Comments

In this manuscript to authors describe an experimental study aimed at determining the branching ratio for the formation of products from bicyclic peroxy radicals (BPR) by making direct measurements of BPR using an iodide chemical ionization mass spectrometer. The measurement methods were optimized to detect BPR at concentrations of ~1 ppt, and various standards and reactions of toluene and xylene with OH radicals under variable NO and HO2 conditions were studied. The results were interpreted using a kinetic model to determine RO2 concentrations and then branching ratios for BPR formation are compared to those determined by others from product yield measurements.

The experiments were very well done, and the overall technical quality of the paper is excellent. The data analysis was also carefully conducted, and the interpretation of the results are reasonable. This is impressive work and opens the possibility for future studies of RO2 radicals that are key reactive intermediates in the atmospheric oxidation of volatile organic compounds. I recommend publication in ACP after the following minor comments are addressed.

Specific Comments

1. These reactions are known to form secondary organic aerosol (SOA), which is not discussed here. How could SOA affect the quantitation of [RO2] by removing gas phase molecular products and RO2 radicals by gas-particle partitioning?
2. What are the lifetimes of initially formed cyclic peroxy radicals (CPR) with regards to ring closure and O2 addition to form BPR and how does this compare to the experiment timescale? How do you know you are measuring all BPR that will be formed in the reaction, and if not, what are the consequences?
3. 4: This equation does not include losses by RO2 + RO2 reactions. Berndt et al. 2018 measured rate constants for self-reactions of BPR near the collision limit, and since [BPR] in Figure 5 are ~ 1 ppb, could these reactions also be sinks?
4. Figure 5: As I understand it, the values of [RO2] in Figure 5 used for calibration were determined from Eq. 6 using literature rate constants; measured [X], [OH], [NO], and [HO2]; and a branching ratio calculated from molecular product yields. Since this approach uses measured product yields, I am not sure it should be called a direct method and the other a product yield method.
5. 15: How are the uncertainties in rate constants and branching ratios taken from the MCM incorporated into the calculated overall uncertainty?
6. Line 404–412: Since the quoted uncertainty in RO2 sensitivities is ~30%, might not all the differences discussed in this section be buried in the errors? Does the comparison between the results here and the product yield method include uncertainties in product yield measurements?

Technical Comments

1. 4: The equation should be ksinks[RO2] = …
2. 13: The second parenthesis should go after [OH].