Lightfastness refers to the permanence of a pigment; some pigments tend to fade, darken, or otherwise change color with prolonged exposure to ultraviolet light, such as when displayed on the wall in a sunny room. The most lightfast a pigment is, the longer it can withstand the sun without changing color. Some pigments can last for hundreds of years. Others – fugitive pigments – tend to fade within months, unless kept away from light (such as in a closed sketchbook) or in carefully controlled museum lighting conditions.

My favorite online pigment experts, such as Bruce MacEvoy and Kim Crick, tend to base their opinions of various pigments in large part on lightfastness. But is this something I really need to care about as a beginner?

(My short answer is no, by the way, but read on.)

Who needs to care about lightfastness?

Lightfastness matters most if your original artwork (not prints) may be displayed in a sunny, bright environment. As an artist, you’ll probably care about lightfastness when:

Selling original paintings. Buyers have an expectation that they can display their art wherever they want and that it should last a good long time.
Gifting original paintings, for the same reason; while giftees are less likely to be irate if your work doesn’t last, they may be sad, and you don’t want to have to give them a laundry list of instructions on how to care for a gift.
Displaying original paintings in the home, especially work you want to last, and especially if you live in a sunny home/part of the world.

However, I find that in most circumstances I find myself in as a beginner/intermediate hobbyist, lightfastness doesn’t matter at all. You don’t need to care about lightfastness when:

Making sketchbook pieces, because sketchbooks tend to remain closed, shielding your work from the sun, until you briefly open them to look at the painting.
Making practice pieces that you immediately scrap (or, at best, display for a few months and then scrap.)
Making pieces to photograph i.e. for posting on Instagram (the original may degrade, but the photo will not).
Selling, gifting, or displaying prints; again, the reproductions will not fade as long as the inks and dyes used to make them are lightfast (probably out of your hands, but companies that make prints should have a handle on this). Note that photographs and prints may not represent colors exactly as they are on your original.

With that said, there’s something to be said for choosing lightfast colors for your palette even when you’re just starting out, especially if you’re the kind of person who tends to stick with the same colors you’re used to. Even if you’re a novice who doesn’t save their work now, you may eventually be in the position of selling or giving your work. Why fall in love with a color you won’t be able to use later?

I tend not to prioritize lightfastness, but I will use it as a tie-breaker when choosing between similar colors. Lightfast pigments come in a range of colors, don’t necessarily cost or more, and aren’t necessarily more toxic; if I have a wealth of choices, and there is no downside to choosing a more lightfast option, why not?

Doing your own lightfastness tests

Every source recommends you do your own lightfast tests rather than relying on others’, as brands can vary, and lighting conditions can vary. I am guilty of mostly relying on other peoples’ since I’m too lazy to do my own, but the process is actually simple. I’ve done it once, and I’m due to do a second round with my new paints.

Paint out a swatch of the color you want to test – a gradient from masstone to dilute is useful, since lightfastness may vary in tints.
Cut the swatch in half the long way, so that each half contains the entire gradient.
Mark each half with the name of the paint, including brand, as well as today’s date. (You can write it on the back.)
Stash one half in a closed sketchbook or other dark place.
Hang the other half in a sunny window.
Wait at least 1 month (if you live in a bright, sunny area or it’s summer), or at least 3 months (if you live in a cloudier area or it’s winter.)
Take down the window swatch and compare it with the stashed swatch.

Lightfast colors should look just the same, where less lightfast colors may have faded.

Lightfast Alternatives to Fugitive Pigments

This list is intended to help you replace fugitive or middling-lightfastness colors from your palette and replace them with lightfast alternatives. Note this is mostly based on research rather than my own lightfast tests. Some paints considered lightfast may still fail in certain brands, and some paints considered fugitive may work perfectly find for you in your environment.

If you prefer more of a “whitelist” approach, see Kim Crick’s list of lightfast pigments.

Aureolin/Cobalt Yellow (PY40)

This is such a famously fugitive color that I have not even tried it! But there are so many great primary yellows to choose from that I don’t feel I’m missing out. From what I can tell, these are the closest yellows to genuine aureolin, a transparent primary-to-cool yellow hat’s a bit warmer in masstone but lemony in dilute:

Quinophthalone Yellow (PY138) has the same quality of appearing like a primary yellow in masstone and being very cool and bright in dilute.
Nickel Azo Yellow (PY150) also has this quality, but its masstone is more of an earthy, ochre type color rather than primary yellow.
Azo Yellow (PY151) has very similar properties though I find it appears more like a cool yellow throughout. Schmincke calls its PY151 ‘Aureolin Hue’!
Holbein also has an Aureolin hue, which is a mix of 3 yellow (PY175, PY154, PY150). This may provide the best of all worlds as it moves from a strong primary yellow in masstone to a cool, transparent lemon.
For a straight-up cool lemon yellow, I love either transparent/semi-transparent Lemon Yellow (PY175) or the more opaque Hansa Yellow Light (PY3). FWIW, I think these two are roughly the same hue, but artist Hilary Page adamantly insists that PY175 is the perfect replacement for genuine Aureolin, and PY3 is too cool.

Brilliant Hansa Yellow (PY74)

I really enjoy this bold yellow, which is a bit warmer than most primary yellows. Kim Crick finds it fugitive in tints, so it might be safer to stick to primary yellows such as:

Pure Yellow (PY154)
Hansa Yellow Medium (PY97)
Cadmium Yellow Medium (PY35) [toxic]

These generally don’t have quite the warmness of PY74, but mixing with a deep yellow (see alternatives below) should get you there.

Indian Yellow (PY83)

This is another great orangey yellow that I truly love, but it fades quite a bit in tints. These deep yellow alternatives are more orangey, but mixing in some primary yellow should get a nice hue:

Isoindolinone Yellow Deep (PY110)
Hansa Yellow Deep (PY65)
Cadmium Yellow Deep (PY35) [toxic]

Scarlet Lake (PR188), Geranium Red (PR242)

Marketed as being “very good” in lightfastness, but Kim Crick finds them both fugitive in tints. These ones are a judgment call rather than an auto-no, but in both cases, the color is really similar to Pyrrol Scarlet (PR255), so maybe an easy replacer.

Naphthol Red (PR112), Naphthol Red (PR170), Anthraquinone Red (PR177)

Kim Crick finds these ones fugitive in tints. I’ve tried all these pigments and didn’t find them to be especially different from other middle reds, which is good news because it means you have plenty of options to replace it. Take your pick:

Pyrrol Red (PR254)
Perylene Red (PR178)
Cadmium Red (PR108) [toxic]
Red versions of PV19, e.g. Alizarin Crimson Quinacridone

Disazo Lake Rose (PR60)

I’ve only ever seen this offered as Holbein’s Geranium gouache. Because it is more common to use gouache only in masstone, it may not suffer from the same lightfastness problems as watercolor, which tend to show up in tints. For example, PY1 yellow is uncommon in watercolor because its dilute end is prone to fading, but it’s a common primary yellow in gouache. However, since Bruce MacEvoy says that PR60 loses saturation in masstone and tint, I think it may be better to avoid even in gouache.

I find to be quite a beautiful color, a rosy coral red that’s hard to match exactly. Winsor & Newton’s Permanent Rose (PV19) is somewhat close, or a mix of Holbein’s Primary Magenta (PR122) and Schmincke’s Vermilion Tone (PR255).

In watercolor, Quin Coral (PR209) is my favorite color with the magical quality of being simultaneously scarlet red and pink, though I find it too transparent to make a good gouache.

Alizarin Crimson (PR83)

Another one I haven’t tried because it’s so well-known for being fugitive. Genuine PR83 continues to be loved by its adherents, but there are some colors that appear to me to have nearly the same hue:

Crimsons made from PV19, such as Da Vinci’s Alizarin Crimson Quinacridone. These may be a bit pinker than PR83, and tend to be fairly similar to other PV19 palette staples (such as Quin Rose).
PR176 reds, such as Daniel Smith’s Carmine. This is a really similar color to Alizarin Crimson Quinacridone, but slightly warmer; I think this is probably the closest match to PR83. However, while less fugitive than PR83, this is not the most lightfast pigment, so you may want to use a different option if lightfastness is your concern.
I also really like crimsons made from PR264 (Pyrrol Crimson), which are a bit warmer and distinguish themselves more from Quin Rose, but Bruce MacEvoy considers PR264 to be “near the bottom end of acceptable lightfastness” so, again, this may not be the best color to choose on that basis.
Some permanent Alizarin Crimson alternatives are made from PR179 (Perylene Maroon); I think this tends to be on the dull side for this color, but may work for you if you want a deeper, warmer crimson. MacEvoy suggests mixing PR179 with PV19 quin rose or PR122 quin magenta to get the best of both words – a dull red and a violet red.

Opera Pink (PR122, BV10)

The word “opera” usually denotes a color with a fluorescent additive, even if the BV10 (or BV1, or similar) is not listed on the pigment label. Opera colors are usually pink, but some companies such as Schmincke also make opera purples.

It is the fluorescent additive that is fugitive, not the PR122 pigment, which is a lightfast magenta. However, it’s also the fluorescent additive that gives Opera Pink its unique, neon glow. There is no lightfast color that looks quite the same. If you don’t want to hitch your wagon to a fugitive paint, your best option is to become one of those people who finds the fluorescence of Opera Pink garish.

By using context (muting down the rest of your colors), you can make lightfast pinks pop without the benefit of hit-you-in-the-face fluorescent. My favorite contextually-glowy pinks are Quin Rose, Quin Coral, or a mix of the two.

Indigo (PB66 or NB1)

A fugitive dye, this is rarely used for colors called “Indigo” with a few exceptions; Schmincke and MaimeriBlu both use it for their Indigo, and MaimeriBlu uses it for their Payne’s Grey as well. A beautiful and interesting color, but go for the more typical mixed Indigos (usually using Phthalo Blue and black) for more lightfastness.

Prussian Blue (PB27)

Kim Crick is adamant that Prussian Blue is a “fugitive pigment often mislabeled as lightfast in watercolors,” and that despite the marketing (it is often labeled as Lightfastness I), “it’s chemically impossible for PB27 to be totally stable” because of the way the sun bleaches the iron oxide that it’s made of. Unlike colors that simply fade, Prussian Blue tends to darken and mute, and often it will actually recover its vibrance if given time to “rest” in darkness. This is a cool feature, but it probably does you no good if you intend to keep the painting out in a sunny location… and it’s probably not something you want to explain to buyers if you’re selling artwork.

I find Prussian Blue very beautiful, and I don’t know of another pigment with the exact same hue or properties, but here are some options:

Indanthrone Blue (PB60) is another gorgeous blue with a similar quality of being muted, moody, and wide-ranging in values. All the PB60s I tried were more violet-toned than all the Prussian Blues I tried, but some PB60s are more violet-toned than others. (I typically go for a more violet-toned PB60, such as Daniel Smith or Schmincke’s Delft Blue, but that’s not what you want when approximating Prussian.) For more of a middle to almost green-toned blue, go for Da Vinci’s or Winsor & Newton’s Indanthrene Blue, or Schmincke’s Dark Blue.
Phthalo Blue Green Shade (PB15:3) has a similar quality of extreme green bias, but is much more saturated than Prussian Blue. However, it’s wide-ranging so you can still get a dark color. The best color match to Prussian Blue I have found comes from adding a bit of red to Phthalo Blue Green Shade or Phthalo Turquoise. Choose a middle or cool red to avoid changing the hue to gray, as a scarlet will do.
Phthalo Blue Red Shade (PB15:1 or PB15:6) is a bit more muted than Phthalo Blue Green Shade (PB15:3); it’s still much brighter than Prussian Blue, but may make it easier to close the gap.

Of course, if you love a fugitive color, go ahead and use it! This is especially true if you work in a sketchbook, or if you don’t care if your work lasts for the ages. It’s also possible to work around lightfastness concerns with UV-blocking frames and so forth. But if you’re hoping to make lasting work (and you can’t control the environments where they’ll be displayed), or you’re building a palette and looking for a basis on which to choose some paints over others, gravitating toward lightfast paints may make your life easier.