DIY Watercolor Lightfastness Tests

Just like colored newspaper hung in a window, some watercolor pigments fade when exposed to light. Sometimes, that fading is on such a long timescale that it’s virtually fadeproof from the point of view of anyone living now (a paint that can withstand direct sunlight for 100+ years would be considered “lightfast”). Other times, the fading may occur in a matter of months (a “fugitive” paint), especially if the painting is exposed to direct sunlight, like being hung in a sunny room. By contrast, even fugitive paint colors will retain their vibrancy if closed in a sketchbook.

So how do you know which paints are lightfast and which are fugitive, and which are somewhere in between?

Watercolor paint companies offer lightfastness ratings of their own paints, but they can be difficult to compare. Between companies, they might have conflicting standards (e.g. companies that rate on different scales, or have different definitions of “good.”) And, of course, they’re motivated to make themselves look as good as possible. You might find that the same pigment fares better or worse depending on the brand, or that your paintings fade more or less than another artists’ depending on the conditions in which you display them.

That’s why online experimenters who write a lot about lightfastness, including Bruce MacEvoy and Kim Crick, suggest that you do your own lightfastness tests. So, I did a couple!


In January 2022, I painted gradient swatches of four colors. Then, I snipped them in half, closing one half into a sketchbook, and displaying the other half in my southwest-facing window, where they were subjected to direct afternoon sunlight. They remained displayed for six months: until July 22, 2022. Where I live, in Massachusetts, USA, this accounts for roughly 1300 hours of sun.


Mission Gold – Brilliant Opera (PR122 + BV10)

Lightfastness tests for Mission Gold - Opera Pink
Left: Closed sketchbook swatch. Right: Sunny window swatch.

(Note: I wrote ‘Opera Pink’ on the swatches but the correct name of the color is Brilliant Opera.)

I expected this notoriously fugitive color to perform the worst, and it did. The BV10 fluorescent pigment completely disappears when exposed to enough light. But I wanted to see how it degraded. As I expected, it totally lost its neon-ness, fading to a standard magenta (that would be the PR122 component). I was surprised by how much lighter it became overall. You need to get the color quite intense before it fades to magenta instead of to super pale pink/white.

Holbein – Marine Blue (PB16)

Lightfastness test for Holbein's Marine Blue (PB16)
Left: Closed sketchbook swatch. Right: Sunny window swatch.

I’d heard conflicting things about the lightfastness of the PB16 pigment, so I performed this test as a genuine “idk what will happen” experiment, using my favorite version of PB16: Marine Blue by Holbein. I was pleased to find very little difference between the two swatches. There is a slight bit of fading in the very dilute end but I find the masstone virtually identical, unaffected by the onslaught of 6 months of sun.

Holbein – Prussian Blue (PB27)

Lightfastness tests for Holbein's Prussian Blue (PB27)
Left: Closed sketchbook swatch. Right: Sunny window swatch.

Prussian Blue didn’t fare as well as Marine Blue. There was considerable fading in the midtone to dilute end, although the masstone looks pretty similar. Also, interestingly, it seems the color became not only lighter but more yellow-toned.

Holbein – Burnt Sienna (PBr7)

Lightfastness tests for Holbein's Burnt Sienna (PBr7)
Left: Closed sketchbook swatch. Right: Sunny window swatch.

This one was a surprise! I didn’t expect this one to change. It was sort of a control. I’ve never heard of PBr7 being anything but lightfast, so I found it shocking that Holbein’s Burnt Sienna changed so much when exposed to light. It didn’t get lighter so much as it changed hue: becoming far less orange and more brown/peach.

I’d been holding onto this color as a sort of backup for Transparent Red Oxide (PR101), but the bizarre lightfastness results compared to my general non-excitement for the color led me to pass this on to another artist. That said, since I haven’t tested TRO, I can’t confirm (yet) that it does better!


This was a sort of random assortment of colors, but I was testing the methodology. It turns out to be quite an easy experiment that I can easily run again now that I’m more firmly decided on my palette colors!

Some suggestions for success:

  1. Write down the date on which you began the experiment on the back of the swatches.
  2. Tape the window swatches to a backboard (not across the paint itself, but over the label or using double-stick tape on back) to keep them organized and prevent them from falling off the window and onto the ground at regular intervals. (Avoid taping the window for six months: it will be hell to get the sticky stuff off.)
  3. Remember where you stashed the protected swatches. I meant to end this experiment after four months, but had to continue it because I couldn’t find the comparison swatches!
  4. Write a calendar reminder to yourself, including the info about where the comparison swatches are, when you want to conclude the experiment.

Happy testing!