Short answer
A pigment colours by absorbing part of the light and returning the rest; a structural colour comes from transparent material arranged at the scale of light's wavelength, which redirects some wavelengths by interference or scattering. The practical tests are angle, transmitted light and what happens when the air in the tissue is replaced by liquid — but many colours are both at once, and several structural colours do not shift with angle at all.
A pigment molecule absorbs photons in a band of wavelengths and converts their energy to heat or chemistry. What is not absorbed is scattered back by the tissue around it, and that remainder is the colour seen. Remove the pigment and the colour goes; grind the tissue and the colour stays, because the molecule is still there. A structural colour involves no absorbing molecule in the visible range. Chitin, keratin, cellulose, guanine and collagen are close to colourless, but when they are laid down in layers, lattices or sponges with a spacing of a few hundred nanometres, light reflected from each interface adds up for some wavelengths and cancels for others. The colour belongs to the geometry and the refractive-index contrast, and it disappears when either is destroyed.
Tilt it. A flat multilayer shifts towards shorter wavelengths as the angle grows, so a green beetle wing case goes blue at a glancing angle; a pigment does not move. Look through it. In transmitted light a structural blue wing shows whatever the structure did not reflect plus any pigment beneath, so a Morpho wing held to the light is brown. Wet it. A structure depends on the contrast between solid and air; filling the air spaces with a liquid of similar refractive index weakens or removes the colour, and it returns on drying, while a pigment is unaffected. None of the three is decisive alone. The third is the most diagnostic, and the first is the one most often over-trusted.
The angle test fails in one direction only. A colour that shifts with angle is structural; a colour that does not shift may still be structural.
Of the 24 structural-colour records in the Colourwise dataset, 16 change strongly or moderately with angle and the remainder change weakly or not at all. The steady ones are steady for three different reasons. A kingfisher's blue barb is a sponge with only short-range order, so it has no preferred direction. A weevil's green scale is a true crystal broken into small domains that point in different directions, so the directional colours of the domains average out. A white beetle scale has no order at all and scatters every wavelength in every direction. The practical consequence is that 'matt and steady' cannot be read as 'pigment' — the blue of a jay feather is the textbook example.
A structural colour needs something dark behind it, or the unreflected light would scatter back and wash it out; in feathers and butterfly scales that backing is usually melanin, and in the fruit of Viburnum tinus it is a layer of anthocyanin. Green in amphibians, reptiles and many birds is typically a structural blue seen through a yellow pigment, which is why a parrot or a frog lacking the yellow looks blue. The kingfisher uses pigment granules for its orange breast and a keratin sponge for its back on feathers a few centimetres apart. Asking 'pigment or structure?' of a whole animal is usually the wrong question; it has to be asked of each patch, and often of each layer within a patch.
Three kinds of biological colour sit outside the dichotomy. Fluorescent proteins and porphyrins absorb short-wavelength light and re-emit it at a longer wavelength, which is neither simple absorption nor redirection. Bioluminescence produces light chemically and needs no illumination. And chromatophore systems are cells or organs that move pigment or tune a structure, so the colour at any moment depends on the animal's state. The dataset therefore files each record under one of four families — pigmentary, structural, light-emitting and cellular control — and records a second contributing mechanism where the source describes one.
| Family | Records | Strong or moderate change with angle | No change with angle | Mechanisms in the dataset | Examples |
|---|---|---|---|---|---|
| Pigmentary | 12 | 0 | 12 | Pigment (selective absorption) | Peppered moth, dark form; Greater flamingo; House finch, male; Red siskin and red-factor canary |
| Structural | 24 | 16 | 2 | Multilayer reflector; Photonic crystal (2D or 3D); Incoherent scattering; Diffraction grating; Coherent scattering, quasi-ordered; Thin-film interference | Morpho butterflies; Green hairstreak; Japanese jewel beetle; Asian jewel beetle |
| Light re-emitted or produced | 7 | 0 | 7 | Fluorescence; Bioluminescence | Reef-building corals; Crystal jelly; Bioluminescent dinoflagellates; Fireflies |
| Cells that move pigment | 2 | 0 | 2 | Chromatophore system | Fish and amphibians with rapid colour change; Common cuttlefish |
| Organism | Structure | Mechanism | Made from | Change with angle |
|---|---|---|---|---|
| Greater flamingo (Phoenicopterus roseus) | Plumage, and preen-gland secretion spread over it | Pigment (selective absorption) | Carotenoids obtained from food, deposited in feathers and also present in the preen-gland secretion the bird rubs on. | None. Absorption colour; no change with angle. |
| Turacos (Family Musophagidae) | Wing flight feathers | Pigment (selective absorption) | Turacin, a copper complex of a uroporphyrin. | None. Absorption colour; no change with angle. |
| Creeping buttercup (Ranunculus repens) | Petal epidermis over a starch layer | Thin-film interference | A smooth, partly transparent, pigment-bearing epidermal layer that reflects like a mirror, backed by a starch layer that scatters yellow diffusely. | Strong. The gloss is highly directional; the yellow beneath is diffuse. |
| Japanese jewel beetle (Chrysochroa fulgidissima) | Wing cases (elytra), epicuticle | Multilayer reflector | Epicuticle stacks of about 16 layers (green areas) and 12 layers (purple areas) with refractive indices between roughly 1.6 and 1.7. | Strong. Both bands move to shorter wavelengths as the light becomes more oblique, and the reflection becomes strongly polarised. |
| Rainbow peacock spiders (Maratus robinsoni, M. chrysomelas) | Abdominal scales about 40 by 10 micrometres | Diffraction grating | Two-dimensional nanogratings on curved, three-dimensional scale surfaces. | Strong. Small movements sweep the reflected colour through the spectrum. |
| Green hairstreak (Callophrys rubi) | Ventral wing scales | Photonic crystal (2D or 3D) | A single-network gyroid of chitin and air: a three-dimensional photonic crystal grown on a folded membrane template. | Weak. Each scale holds many small crystal domains in different orientations, which averages out the direction-dependence of any one domain. |
| Common kingfisher (Alcedo atthis) | Back and tail feather barbs | Coherent scattering, quasi-ordered | Spongy keratin-and-air nanostructure of slightly different dimensions in cyan and blue barbs, inside a cortex a few micrometres thick; the orange barbs hold pigment granules instead. | Weak. Scatterometry shows some angle-dependence, increasing towards shorter wavelengths; far less than a multilayer. |
| White scarab beetle (Cyphochilus spp.) | Body scales about 5 micrometres thick | Incoherent scattering | A disordered, anisotropic network of chitin filaments and air filling each scale; no pigment. | None. Diffuse, broadband reflection that looks alike from all directions. |
| Reef-building corals (Anthozoa) | Host tissue | Fluorescence | GFP-like fluorescent proteins made by the coral itself. | None. Emitted in all directions. |
| Fireflies (Family Lampyridae) | Abdominal light organ | Bioluminescence | A luciferin–luciferase reaction under nervous control of timing. | None. Self-luminous. |
| Common cuttlefish (Sepia officinalis) | Skin chromatophore organs | Chromatophore system | Each organ is an elastic pigment sac stretched open by radial muscles under direct nerve control and closed by its own elasticity. | None. The pigment sacs are angle-independent; reflecting cells beneath them are not. |
Each statement is labelled by kind — established fact, a standard’s requirement, observed market data, a convention, or Colourwise’s own interpretation or analysis — with the strength of the evidence behind it.
FactStrong evidence
Across vertebrates, blue is almost always a structural colour produced by coherent or incoherent scattering; a comparative review found a single known case of a blue pigment held in a pigment cell.
Caveat: Concerns vertebrates; invertebrates include a few more blue pigments.
FactStrong evidence
The quasi-ordered nanostructure of feather barbs changes colour with the angle between illumination and observation under a single directional beam, but is invariant with viewing angle under light arriving from all directions.
Source: How noniridescent colors are generated by quasi-ordered structures of bird feathers
FactStrong evidence
For amphibians, reptiles and birds, green typically results from the scatter of blue wavelengths combined with yellow pigmentation rather than from a green pigment.
Caveat: Turacos are a known exception among birds, with a copper-based green pigment.
Colourwise analysisModerate evidence
In the Colourwise biological-colour dataset, 16 of 24 structural-colour records change strongly or moderately with angle, against none of 12 pigmentary records.
Based on: Each organism record in the Colourwise biological-colour dataset is counted once, under its primary mechanism and its function-evidence grade. Non-living analogies are excluded. The records were chosen to illustrate mechanisms, not sampled from nature, so the counts describe this dataset and the studies it cites — not how common each mechanism or each kind of evidence is among living things.
Caveat: A count over records chosen to illustrate mechanisms. It shows that angle-steady structural colour exists and is not rare among studied cases; it is not a frequency in nature.
Source: Physics of structural colors; How noniridescent colors are generated by quasi-ordered structures of bird feathers; Discovery of a diamond-based photonic crystal structure in beetle scales
Colourwise interpretationStrong evidence
Change with angle is a sufficient but not a necessary sign of structural colour, so the absence of iridescence should never be used on its own to conclude that a colour is pigmentary.
Based on: Follows from the measured angle-steady structural colours of spongy feather barbs, mammalian collagen arrays and multi-domain photonic crystals recorded in the dataset.
Source: How noniridescent colors are generated by quasi-ordered structures of bird feathers; Structural colouration of mammalian skin: convergent evolution of coherently scattering dermal collagen arrays
Reviewed 6 October 2026. Colourwise summarises its sources in its own words and does not reproduce standards text or proprietary colour data. Spotted an error? Tell us.