Dart width is the distance the dart needs to be opened at the seam (the shoulder seam in this case) in order to allow for the correct shaping at the bust. It’s an arbitrary amount you’ll find in my women’s sizing table for ready-to-wear sizing, but for custom made garments you’ll need a formula.

Simply explained, the bigger the bust compared to the waist, obviously the more dart angle required for a good fit. So it’s a mathematical function of angle and distance, where angle is based on bust/waist ratio and width is based on dart length … and seeing as we can easily measure all these it’s reasonably straight forward to create a formula for dart width.
Dart Width = ((Bust – Waist) x (Dart Length)) / (4 x Waist To Bust Point)
For example, taking the measurements for the standard size 10 with the dart positioned at the shoulder we can use the mid shoulder to bust point measurement as dart length and get ….
Dart Width = ((84 – 65) x 22.8)) / (4 x 15.7) = 6.9cm
Now reduce that for 10% negative ease (I ease the darts back a little from 12% but that’s up to you) and you get a dart width of 6.2cm. Remember this is a proportional exercise and the whole block is scaled down so the dart is also scaled down in order to maintain the constant angle.
Now I know there are going to be people jumping up and down saying the result doesn’t match the table … and that’s correct, but the table sizes are ready to wear sizes, and are based on a mix of my measurements, the ISO measurements and a requirement for the dart width to increase in a linear fashion. So as a standardised table (OK there really isn’t such thing as a standard) it’s not going to reflect a custom formula result. Moral of the story is, if you want a made to measure fit then use the formula … if you’re creating ready to wear size groups you are going to need a commercial incremental overlap so use the table!.
I also really should modify the women’s sizing table to remove the negative ease from that measurement, it’s just that everyone ignores this part no matter how much I tell them … and if they do accept it they think they need to increase the dart width rather than reduce it. So I usually specify a fixed amount, hide the ease in it and just tell people to open the dart by that fixed amount as if it’s a constant value. I’d rather not do this, but ready to wear designers aren’t known for taking time to think things through. Of course custom wear people will be able to use the above formula.

Ahh! I get it! Thanks so much, this will be very handy. After at first being confused, I realised it was based on the tan-theta opposite/adjacent ratio – the same ratio will apply (pretty much) between quarter bust/waist difference and waist-to-bust, as will apply between dart width and bust-to-shoulder.
The patternmaking books I’ve come across so far seem to omit this flexible way of calculating bust dart.
Is it correct to you opinion to draft without negative ease and then scale it down on the CAD X/Y wise as much as each fabric require?
As this is a theory for before you apply negative ease then it should work for shirts before you add ease.
Hi there
Ifound this quite interesting as I work in the tailor made area with ladies shirts and have found the fit isnt the best even though we have added darts – would this formula work for shirts also?