This is not easy to determine! It depends on the type of device you are using – IPL, diode, alexandrite, Nd:YAG laser. It depends on the depth of the follicles. It depends on the hair colour.
I built a computer model to try to calculate these fluences. It considers all of the above, but today I will just consider an IPL treating very dark hair, as a starting point.
First, we should consider what we are actually trying to do…
To effectively kill hair follicles, we need to essentially ‘cook’ the germ/stem cells, irreversibly. To achieve that, we must apply a certain amount of energy over a certain time – just like any cooking!
So, we need to find the minimum amount of heat (light) needed to cook these cells properly.
My model allows me to apply different fluences over different pulsewidths to hair follicles at various depths.
Here is what I found (for IPLs):
The depths I used here are for the bulge, where the stem/germ cells reside on the follicle wall surfaces.
These fluences assume a pulsewidth between 2 and 12 ms. Shorter, and longer, pulsewidths will not work so well – the cooking is less efficient! I have also assumed a wavelength range between 600 and 1200 nm in these calculations. A smaller range will require higher fluences.
Obviously, deeper follicles (bikini, armpits) require more fluence at the surface to ensure sufficient light gets deep enough to heat them. This means that more pre-cooling is also required, to protect the epidermis.
Body Area | Threshold fluence (J/cm^2) |
Upper Lip | 20 |
Chin/Arms | 28 |
Arm pits/legs | 33 |
Bikini | 36 |
Interestingly, a small increase in fluence (i.e. 1 J/cm2) above these threshold makes a large difference in terms of cell destruction! But, fluences below these levels will merely destroy some of the stem/germ cells, meaning that finer hairs will likely grow back in the future.
These are my first set of calculations – they may change as I delve deeper into this study. But, I know from clinical experience that the numbers above make sense, for those pulsewidths.
I plan to generate further sets of these fluence/pulsewidth combinations for diode, alexandrite and Nd:YAG lasers in the near future. I will be discussing all this stuff in much more detail in my MasterClasses.
Hope this helps,
Mike.

If your IPL’s shorter Pulse width is programmed between 9-39ms (not 2 -12 ms) with a wave length of 770-1100nm, would your threshold fluence yield good results?
Can you use a lower wavelength filter? Below 770nm?!
No filters. There are 3 different programs that can be selected to tailor wavelength and pulse width.
Which device is this? I’ve not heard of this approach?
Cutera Prowave 770