The FWHM is calculated by Skyline looking at the chromatogram for a particular transition. You can see those FWHM values in the document grid by looking at the column:
Proteins > Peptides > Precursors > Transitions > Transition Results > Fwhm

There is also a column called "Max Fwhm" which can be found at:
Proteins > Peptides > Precursors > Precursor Results > Max Fwhm
This "Max Fwhm" column contains the maximum Fwhm value across all of the transitions for the particular Precursor in a particular Replicate.
This Max Fwhm column displays its value as a simple number, such as "0.36".

There is another column which is also called "Max Fwhm" which can be found on the "Precursor Result Summary" at:
Proteins > Peptides > Precursors > Precursor Results Summary > Max Fwhm
This precursor result summary Max Fwhm gets its value by looking at the "Max Fwhm" values for a particular Precursor across all of the Replicates.
This Max Fwhm column displays its value as the average of the "Max Fwhm" values across all of the replicates, with a "+/-" indicating the standard deviation of those same "Max Fwhm" values across all of the replicates.

The Precursor Result Summary Max Fwhm has some child columns which you could also add to your report:
"Mean Max Fwhm"
"Stdev Max Fwhm"
"Cv Max Fwhm"

I am not sure that I understand your question. Can you send us a screenshot of what you are seeing, and/or can you send us your Skyline document and the report that you are using?

You can make it so that your report definition is included in your Skyline document. To do this, go to:
Settings > Document Settings > Reports
and check the checkbox next to the report that you have questions about.

After you have done that, you can go to:
File > Share
and create a .zip file containing your Skyline document and supporting files including extracted chromatograms.

If that .zip file is less than 50MB you can attach it to this support request.
You can upload larger files here:
https://skyline.ms/files.url

-- Nick

Firstly, the FWHM was exported as follows. Proteins > Peptides > Precursors > Precursor Results Summary > Max Fwhm
Secondly, I extracted some target peptides from DIA files by skyline. The attached peptide precursor was extracted from two files (right one shows a good MS1 peak; left one show a bad MS1 peak). The FWHM exported from the left file is 5 times longer than the right file. The FWHM of the left one must be wrong. How to set the skyline to prevent this issue. In my opinion, FWHM of the left file would be zero or NA instead of a wrong number.

The "Max FWHM" value that you are seeing is the FWHM of one of the transition chromatograms.
I would recommend that you look at the FWHM values for each of the transitions, and figure out which chromatogram is responsible for the Max FWHM value that you are seeing.
That is, you should show the Results Grid using the menu item:
View > Other Grids > Results Grid
and then you should select each of the transitions in the Targets tree, one at a time. The results grid will display the FWHM values for the selected transition across all of the replicates.

The FWHM value is determined by finding the two points farthest from each other within the integration boundaries where the background-subtracted intensity on the chromatogram is at least half of the background-subtracted height at the apex of the peak. Here is a page which shows how Skyline calculates the background level of an integrated peak:
https://skyline.ms/wiki/home/software/Skyline/page.view?name=tip_peak_calc

When you have jagged peaks with multiple apexes like you have in the chromatograms on the left, I am not sure that FWHM will be a useful number for you.
-- Nick

-Nick
When you have jagged peaks with multiple apexes like you have in the chromatograms on the left, I am not sure that FWHM will be a useful number for you.

-Winnie
In my opinion, we should evaluate the quality of the peak before doing the FWHM calculation. If the peak like the left one, it would be better to output a NA instead of a wrong number.

Thanks again.

I see the point that "Max FWHM" may not be all that useful in this case. It clearly has the expectation that all peaks are well-formed. Maybe it should have been or should be the FWHM of the peak with the maximum Area value.

I am not quite sure what criteria you would use to decide that a peak is not well-formed enough to be given an FWHM value. Can you articulate your thoughts on that? I was thinking maybe based on Area:Background ratio, but then remembered that Agilent artificially sets its background at 42. So, a ratio like that is not likely to work equally well across all instrument vendors.