The short answer: the color rendering index scores a bulb from 0 to 100 on how accurately it shows colours compared with daylight, and it is the reason a room with perfectly good lighting can still look flat. Most inexpensive LEDs are CRI 80, better ones are 90 or above, and the step between them costs very little. But the number is an average of only eight test colours, and the deep red sample called R9 is deliberately left out of it. That is why two bulbs can both claim CRI 90 while one makes a bowl of tomatoes look rich and the other makes it look grey. If you want a shortcut: buy one bulb before you buy eight, and look at something red under it.
Disclosure: this post contains affiliate links, and as an Amazon Associate I earn from qualifying purchases. If you buy through them I may earn a small commission at no extra cost to you. I only recommend things I would use in my own home.
What the number is actually measuring
Most explanations of CRI are written by lighting manufacturers for other lighting professionals, and they are accurate and almost useless if what you want is a kitchen where food looks appetising. Here is the same thing explained for someone furnishing a room.
Daylight contains everything. Bulbs do not
Sunlight carries every wavelength in roughly even proportions, so every colour it lands on is reflected back accurately. That is why things look right outdoors and why daylight is the reference the whole scale is measured against.
An LED produces light in a different way, by exciting phosphors, and the resulting spectrum has peaks and gaps in it rather than being smooth. Any colour that happens to fall into a gap has almost nothing to reflect, so it comes back muted. The bulb is not dim and the colour temperature can be perfect. The colour simply is not in the light to begin with.
How the score is calculated
Eight standardised test colours, labelled R1 to R8, are compared under the bulb and under a reference light at the same colour temperature. Each gets a score from 0 to 100, and the CRI you see on the box is the average of those eight.
Two things about that method matter to you. The eight samples are all pastel, low to moderate saturation shades, so the test is gentler than the objects in a real room. And the average hides the individual scores completely, which is the weakness the next section is about.
Numbers worth knowing
Where the common light sources actually land. Daylight is the reference and scores approximately 100. Incandescent and halogen bulbs sit above 95, which is the one thing they were genuinely better at than LEDs. Fluorescent tubes typically manage 70 to 80, which is why an old kitchen strip light made everything look tired. LEDs vary enormously, from below 70 in the cheapest bulk packs to above 95 in bulbs sold specifically on colour quality.
The standard itself dates from 1937, which is worth knowing because it was designed for the light sources of that era. The lighting industry has since developed a newer measure, TM-30, that tests 99 colours rather than 8, but consumer packaging has not caught up and CRI is still what you will find on a box in a shop.
So the practical summary is that CRI answers one question well and hides another. It tells you roughly how faithful a bulb is across ordinary pastel shades. It tells you almost nothing about how it handles a saturated red, and that is the failure people actually notice.
Find the CRI you need in 3 taps
Where the light goes, what you look at under it, and how much you want to spend. Those three settle it.
R9: the number they leave out
This is the part that explains why CRI can be technically correct and practically misleading, and it is why two bulbs both stamped CRI 90 can look completely different in the same kitchen.
The average excludes the hardest colour
CRI averages eight test colours, R1 to R8, all of them pastel and moderately saturated. There are further test colours beyond those eight, and R9 is the deep saturated red. It is calculated, it is reported separately by manufacturers who care, and it is not part of the headline average at all.
Which produces exactly the situation you would expect. A bulb can score respectably across the eight gentle samples, average out to 80 or 90, and be genuinely poor at red. Nothing about the number on the box reveals it. As one lighting supplier puts it in their own technical documentation, a fixture can score CRI 90 while rendering reds poorly if its R9 is low.
Why red is the one that matters at home
Red is not an arbitrary choice of weak point. It is in almost everything you care about how it looks. Tomatoes, peppers, raw meat and cooked meat. Terracotta, brick and warm wood. Wine. And skin, which has its own separate test value, R13, that depends on the same red content and is also excluded from the average.
So the two rooms where colour rendering fails most visibly, kitchens and bathrooms, are failing on precisely the value the standard leaves out. That is not a coincidence, it is the mechanism.
Numbers worth knowing
R9 is scored on the same 0 to 100 scale but it is mathematically harder to score well on, because it is more sensitive to gaps in the spectrum. So the thresholds are different from CRI: an R9 of 50 or above is considered good, and 90 or above is excellent. A bulb quoting R9 above 50 is telling you something meaningful, which is more than most tell you at all.
The honest caveat for anyone shopping in a normal shop: consumer bulb packaging almost never states R9. It appears on photometric reports and on products sold to photographers, retailers and galleries. So for household buying, treat a published R9 figure as a strong signal that the manufacturer takes colour seriously, and treat its absence as a reason to test one bulb rather than as evidence of a bad bulb.
The test that costs one bulb
Since the number you need is usually not printed anywhere, the practical method is embarrassingly simple and genuinely reliable. Buy one bulb rather than a pack. Fit it in the room it is meant for. Put something strongly red under it: a tomato, a red apple, a glass of wine, your own hand.
Then look at it in daylight and look at it again under the bulb. If the red loses depth and goes slightly brown or grey, the R9 is poor whatever the box claims. That takes five minutes and it is more informative than any specification you can read online.
CRI by light source, and what to expect
Useful for two reasons: it tells you what you are comparing against, and it explains why a room can look worse after a perfectly sensible upgrade.
| Light source | Typical CRI | Red rendering | What it means at home |
|---|---|---|---|
| Daylight | 100 | Perfect | The reference everything else is measured against |
| Incandescent | 95 to 100 | Excellent | The one thing old bulbs did better. Continuous spectrum, no gaps |
| Halogen | 95 to 100 | Excellent | Same continuous spectrum, which is why galleries kept using it |
| LED, budget | 70 to 80 | Often poor | The bulk pack problem. Fine in a cupboard, wrong in a kitchen |
| LED, standard | 80 to 85 | Variable | The default in most shops. Acceptable rather than good |
| LED, high CRI | 90 to 98 | Good to excellent | Costs slightly more and is the best value upgrade in lighting |
| Fluorescent tube | 70 to 80 | Poor | Why old kitchen strip lights made everyone look unwell |
The two incandescent rows explain something people notice and misattribute. If a room felt warm and rich under old bulbs and looks flat since you switched to LED, the colour temperature may be identical and the colour rendering is not. Incandescent light has a continuous spectrum with no gaps in it, because it works by heating a filament until it glows rather than by exciting phosphors. That is a physical difference, not a quality difference between brands.
Two things CRI does not tell you
It says nothing about the colour of the light. A 2700K bulb can be CRI 95 or CRI 75, and so can a 5000K one. The two specifications are completely independent, which is covered in more detail in the guide to the three numbers on a bulb box.
And it says nothing about brightness. That is lumens, and a high CRI bulb at 450 lumens will still leave a room dim.
The links below are affiliate links. As an Amazon Associate I earn from qualifying purchases.
The one to test first: a dimmable CRI 90 bulb
Buy one of these before you buy anything in quantity. That is not a hedge, it is the method: since R9 is almost never published on consumer packaging, a single bulb in the actual room under the actual conditions tells you more than any specification sheet.
This one states CRI 90 plainly, runs at 2700K and 800 lumens, which replaces an old 60 watt bulb, and is dimmable, which matters because the ability to take the light down in the evening is what makes a warm bulb feel restful rather than merely yellow. Rated for 25,000 hours and ENERGY STAR qualified.
One caution from the manufacturer worth repeating: it is not recommended for fully enclosed fixtures, because trapped heat shortens LED life considerably. If your fitting has a sealed glass shade, look for a bulb specifically rated as enclosed-fixture safe.
Worth it or skip: Worth it as a single test bulb regardless of what you eventually buy in bulk, and worth it permanently in a kitchen, a bathroom or a main living space. Skip it for a hallway, a garage or a cupboard, where CRI 80 is genuinely adequate and the saving across a house is real.
Three more, once you know what you want
A CRI 90 six pack, for a whole kitchen. Same colour accuracy and the same 2700K and 800 lumens in a multipack that works out cheaper per bulb. It is non-dimmable, so check your switches before ordering. Buy the single test bulb first if you have not already, because six of something you dislike is a worse outcome than one. See the CRI 90 six pack →
CRI 90+ spotlights, where colour distortion shows most. A narrow beam falling on one object is the least forgiving lighting there is, because there is nothing else in the frame to average against. PAR20 at 500 lumens and 2700K, dimmable. If you have recessed spots over a worktop or aimed at artwork, this is where the CRI difference is most visible. See the PAR20 spotlights →
A standard CRI 80 eight pack, for everywhere else. There is no virtue in buying high CRI for a cupboard, a landing or a garage, and across a whole house the saving is meaningful. Dimmable, 2700K, 800 lumens. Put the good bulbs where you cook and where you look at your own face, and use these for the rest. See the standard eight pack →
Working out the colour temperature and brightness as well? The guide to the three numbers on a bulb box covers Kelvins and lumens alongside CRI, with a room by room table.
Diagnose what looks wrong in 2 taps
The quiz above is for buying. This one is for a room you already live in where something looks off and you cannot name it.
People also ask about the color rendering index
What does CRI mean in lighting?
Colour Rendering Index, a score from 0 to 100 measuring how accurately a light source shows colours compared with natural daylight. Daylight is the reference and scores approximately 100. It is calculated by comparing eight standardised test colours under the bulb against a reference light at the same colour temperature, then averaging the eight scores. Crucially it describes colour accuracy only: it tells you nothing about how bright the light is, which is lumens, and nothing about whether it looks warm or cool, which is Kelvins.
Is CRI 80 good enough?
It depends entirely on the room. In a hallway, a garage, a cupboard or a landing, CRI 80 is genuinely fine and buying better is spending money where nobody will notice. In a kitchen, a bathroom or anywhere you look at people, it is not: 80 is where food starts to look grey and skin starts to look tired. The sensible approach is to split the budget rather than spread it, putting CRI 90 where you cook and where you look in a mirror, and standard bulbs everywhere else.
What is R9 and why does it matter?
R9 is the deep saturated red test colour, and it is not included in the CRI average. That single fact explains most of the confusion in this subject: a bulb can score CRI 90 across eight gentle pastel samples while rendering strong reds poorly, and nothing on the box reveals it. R9 is scored on the same 0 to 100 scale but is harder to score well on, so above 50 is considered good and above 90 excellent. Skin tone has its own excluded value too, R13, which depends on the same red content.
Do LED bulbs have worse CRI than incandescent?
On average yes, and for a physical reason rather than a quality one. Incandescent and halogen produce light by heating a filament, which gives a continuous spectrum with no gaps, so they score 95 to 100 almost automatically. LEDs excite phosphors, producing a spectrum with peaks and gaps, and colour rendering has to be engineered in. Budget LEDs land at 70 to 80 and good ones reach 90 to 98. So the answer is that LEDs vary where incandescent did not, which is why the number is worth checking now and was not before.
Color rendering FAQ
Where do I find the CRI on a bulb?
Usually on the box, often in the specification block on the back rather than the front, and sometimes printed on the bulb itself. Online it is normally in the technical details rather than the headline description. If a listing does not state it anywhere, treat that as a signal: manufacturers who have invested in colour rendering tend to advertise it rather than bury it.
Does high CRI cost much more?
Less than people expect. Comparable A19 bulbs from the same shelf at CRI 80 and CRI 90 differ by a small amount per bulb, and for a kitchen of six fittings the total difference is roughly the price of a takeaway. Set against a bulb life of around 25,000 hours, it is the best value upgrade available in lighting and the one almost nobody makes.
Does dimming change the CRI?
On a well made LED, not meaningfully. Cheaper drivers can shift the spectrum slightly at very low output, and some bulbs go faintly greenish or bluish when dimmed hard, but the effect is small compared with the difference between an 80 and a 90 CRI bulb at full output. If you notice a colour shift when dimming, that is usually a driver quality issue rather than a CRI one.
What is TM-30 and should I care?
A newer measurement developed because CRI, which dates from 1937, was designed for the light sources of that era. TM-30 tests 99 colour samples rather than eight and reports both fidelity and saturation. It is a better tool and it is used in professional specification. For household buying it is not yet relevant, because consumer packaging has not adopted it and you will not find it in a shop.
Do smart bulbs have good colour rendering?
Often worse than a fixed-temperature bulb at the same price, and for a structural reason: the same LEDs have to cover a wide range of colour temperatures, which is harder than optimising for one. Check the CRI figure before buying, and be aware that many colour-changing bulbs quote the figure only for their white mode. If accurate colour matters in that room, a good fixed bulb usually beats a smart one.
Can I tell the CRI just by looking?
Not as a number, but you can tell the difference between two bulbs reliably, which is what actually matters. Put something strongly red under each in the same room and compare, then compare both against daylight. Poor red rendering shows as a shift toward brown or grey rather than as dimness. Five minutes with a tomato is more useful than any specification you can read.
The CRI cheat sheet
What the numbers mean
Below 80: poor. Cheapest bulk packs and old fluorescent tubes.
80 to 85: the ordinary default. Fine in a hallway, wrong in a kitchen.
90 to 98: where colours look right. Costs slightly more.
R9 above 50: good red rendering. Above 90 excellent. Rarely published.
Where to spend
CRI 90: kitchen, bathroom, dining, main living space, anywhere with art.
CRI 80: hallway, landing, garage, cupboard, utility, basement.
Split the budget rather than spreading it. Good bulbs where you cook and where you look at your face.
By light source
Daylight 100 · Incandescent and halogen 95 to 100 · Fluorescent 70 to 80.
LED anywhere from below 70 to 98. This is the only source where you have to check.
The test that beats the spec sheet
Buy one bulb, not eight. Fit it in the actual room.
Put a tomato, a red apple or your own hand under it, then look at the same thing in daylight. If the red goes brown or grey, the R9 is poor whatever the box says.
Free printable
The Weekly Home Reset Checklist
Good light shows a room exactly as it is, which is a reasonable argument for keeping it in order. This printable is a short weekly loop that keeps a minimal home actually minimal, without turning tidying into a project.
Keep reading
- A Bulb Has Three Numbers. Most Guides Explain One
CRI is the third of them. This covers the other two, Kelvins and lumens, with a room by room table for both. - Scandinavian Lighting Guide: Create a Cozy Hygge Glow
Good bulbs in one central fitting still make a flat room. This is how to layer light so the quality shows. - Minimalist Lighting Ideas: How to Illuminate Your Scandinavian Home
Fittings room by room, for when the bulb is sorted and the question becomes where to put it. - Mirror Height Has Two Answers, and a Third Nobody Mentions
Skin tone rendering matters most at a bathroom mirror, and where the mirror sits decides how the light falls on your face. - Minimalist and Scandinavian Design: How to Create a Simple, Cozy, and Stylish Home
The framework underneath all of it, and why light quality sits at the centre of the whole style.
One number, one exclusion, one tomato
Colour rendering index is a genuinely useful specification and it is also the reason two bulbs can carry identical numbers and look nothing alike. The score averages eight gentle pastel test colours and deliberately leaves out R9, the deep saturated red, along with R13, which is skin. Those two are precisely what fails in a kitchen and a bathroom, which is precisely where people notice.
So the practical version is short. CRI 90 or above where you cook, where you wash and where you look at people. CRI 80 in the hallway, the garage and the cupboard, because spending more there is spending money nobody will see. And treat a published R9 figure as a sign the manufacturer takes this seriously, rather than as something you will usually be able to find.
Then do the thing that no specification replaces. Buy one bulb, fit it in the room it is meant for, and put something red under it. Compare that to daylight. If the colour goes brown or grey, the number on the box was telling you the truth about eight pastel samples and nothing about the tomato in your hand.
My kitchen ran on CRI 80 for years and I assumed I simply did not like the room. It cost about fifteen pounds to find out I had been wrong about that.
Leave a Reply