Farnsworth Munsell 100 Hue Test
85 generated colors · 4 trays · browser practice
Not a diagnosis or professional advice. For a color-vision assessment, see an ophthalmologist. Screen colors and lighting affect this exercise.
Tray 1 — Red to yellow
0 of 4 readyArrange a smooth color sequence from Start to End. Read left to right, then continue on the next line. Tap two tiles to swap; or select one and use the move buttons.
Select a tile to move it.
No timer. Take the time you need. Progress lasts until you reload or start a new round.
YOUR COMPLETE RESULT
This round's error score
Total error score · lower means fewer ordering errors · 0 is a perfect sequence
These tray scores describe errors in the four generated color sequences. They do not identify protan, deutan, or tritan vision, or measure the severity of color-vision deficiency.
Every tile's error score and my final order
Generated HSL hues; adjacent-rank scoring with fixed anchors. This screen exercise has no validated clinical norms. The historical bands below belong to the named physical-test method and are not a grade for your eyesight.
PRINTABLE REFERENCE
Score bands: source, date, and limits
For the physical FM100 test, He, Bex & Skerswetat's June 15, 2023 preprint, Rapid measurement and machine learning classification of color vision deficiency, reports these Total Error Score bands in its FM100 results section. This is the cited preprint version, not a claim about clinical validation of this page.
| TES in cited method | Source label | How to read it here |
|---|---|---|
| 0–16 | Superior discrimination | A historical physical-test category, not proof of normal color vision. |
| 17–19 | No band assigned in this source | The cited intervals leave this gap. We do not invent a category. |
| 20–100 | Average discrimination | The source starts this interval at 20, not 16. |
| Over 100 | Low discrimination | A physical-test discrimination label, not a diagnosis or deficiency severity. |
Scope: these bands belong to that physical-test scoring context. They are not universal, age-adjusted norms and must not classify this generated-color practice. The National Research Council's Procedures for Testing Color Vision (1981) describes display-independent physical samples, lighting requirements, and the need for age and illumination norms.
To keep this static comparison table, use your browser's Print command. The print view includes this reference and, after completion, your result. No download is needed.
Reference checked October 4, 2026. Not a diagnosis or professional advice; consult an ophthalmologist for assessment.
How this exercise is scored
The 1981 National Research Council description explains neighboring-cap errors and summing the four boxes. We use the published adjacent-rank formula, with the anchor-aware endpoint variant compared by Viliūnas and colleagues (November 2006). The 2018 paper An Adjusted Error Score Calculation for the Farnsworth-Munsell 100 Hue Test describes the standard adjacent-rank formula and four-tray structure.
Tile error = |tile rank − left neighbor rank| + |tile rank − right neighbor rank| − 2
At each tray's ends, the neighbor is the fixed anchor's rank. Anchors contribute no separate score. Sum all movable tile errors for the tray, then sum the four trays for the total. A correct sequence gives 1 + 1 − 2 = 0 per tile. Swapping two adjacent interior tiles in an otherwise correct tray gives a total of 4. Review the tile details in your result to check the arithmetic.
This page uses 85 movable tiles in trays of 22, 21, 21, and 21, following the published four-tray structure. We generate hues at equal steps of 360/85 degrees around an HSL circle, with saturation 46% and lightness 58%. The surrounding background is neutral. These settings describe our code, not standardized Munsell cap measurements; equal HSL steps are not equal perceptual differences or equal luminance. The colors are generated here, with no physical card images.
The first and last anchors repeat the adjacent hue from the color circle. Rank 0 is the preceding endpoint and rank 86 the following endpoint. This keeps rank differences continuous within each tray; the four tray scores are an exercise breakdown rather than clinical confusion axes. Anchor-aware and traditional cross-box scoring can differ, as the 2006 paper explains.
Check one image with the simulator
Use this exercise to practice ordering colors. To inspect a design, use the separate image simulator and the same six-step review shown in our protanomaly guide:
- Name the decision. Write what the viewer must identify, such as the status needing attention.
- Choose one image. In the simulator, open the exact screen, chart, map, or photo. Keep the same crop.
- Set the view. Choose Protan and record the method and simulation strength. These are model controls, not a measurement of eyesight.
- Find the collision. Compare Original, Simulated, and Difference; record the labels, marks, or boundaries that become hard to distinguish.
- Add a second cue. Edit the source design to add a label, shape, pattern, border, or position cue.
- Repeat the same decision. Open the edited image and repeat with the same crop, method, and strength. Record whether the added cue made the task clear.
This test page has no image input. The simulator processes the image locally in your browser. Its auto-scaled Difference panel is not a score of anyone's vision.
Frequently asked questions
Is this the full Farnsworth 100 hue test?
This Farnsworth 100 hue test inspired exercise includes all four trays and 85 generated movable colors. It is a browser practice task, not the standardized physical instrument. The traditional “100” name does not mean this page has 100 movable tiles.
What is a good Munsell hue test score?
Zero means no ordering errors in this exercise; lower totals mean fewer neighboring-rank errors. Historical physical Munsell hue test bands are printed above with their source and date. They do not establish a good clinical score for an uncalibrated screen.
Can a Munsell 100 hue test diagnose color blindness?
This Munsell 100 hue test inspired page cannot diagnose color blindness or grade its severity. Generated HSL colors, screen settings, room light, viewing angle, and practice affect the task. For an assessment, see an ophthalmologist.
Can I finish the 100 hue test on my phone?
Yes. Tap one tile and then another to swap them, or select a tile and use Move earlier or Move later. Read across each line before continuing on the next. Finish four trays to see every score. No login or email is required.
Is my result saved or uploaded?
Your arrangement is scored in this browser. This page does not send your result to a server or keep a history. Copy your result before reloading if you want to keep it.
Can I use a keyboard?
Tab to a tile and press Enter or Space to select it. Select a second tile to swap, or use the arrow keys to move the selected tile earlier or later. Escape clears the selection. Undo reverses the latest move in the current tray.