Thanks for your attention to the reviewers' comments. There are still some issues and unanswered questions that I would ask you to attend to.

You did not answer Reviewer #1's question about what fraction of OH reactivity is represented by these pyrolysis emissions.

Figure 5 is redundant and does not add to the discussion, it should be eliminated.

It seems clear that NH3 will be lost in the analytical system, but the authors are incorrect that other nitrogen species could be lost too, (line 525 revised text). In fact HCN and CH3CN are quite well transmitted by stainless steel sampling components. but the authors are correct that amines will also be readily lost in their system.

The authors quote the arithmetic mean is given in Table 2, but the Authors' reply says the geometric mean is the most appropriate - why was it not given instead?

The emphasis on peat content is still in the manuscript (line 548), it doesn't seem like these fuels have any peat content. What the authors' did not explain is that the magnitude of nitrogen compound emissions depends on the N-content in fuels, and did not connect their emission measurements with the fuel nitrogen measurements - are there dependencies there that shown up in the data?

Lines 62-66. These two sentences seem at odds with each other.

Line 238. It seems in appropriate to call these ERs discrete since they are averages of a number of aliquots. The sampling scheme is intended to one set of combustion conditions, but the samples are not really discrete.

Lines 312-315. This is a virtual repeat of material in lines 305-308.

Lines 416-418. This material is repetitious, and Figure 5 does not add anything new to the discussion, it can be eliminated and the text tightened up.

Line 421. How does point d) lead to this effect. Doesn't calibration account for this? Or are you saying you are missing entire compounds or classes because they are not detected?

Line 582-583. I suggest rephrasing this, as it is hard to follow all the negatives.