The Verification Venue · a thing every school diagram gets wrong
The Blood That Was Never Blue
Your blood is never blue, not deep in your veins, not for a microsecond, not at any oxygen level. Oxygen-rich blood is bright scarlet; oxygen-poor venous blood is dark maroon. So why do the veins in your wrist look blue? And why is the answer you were told also wrong?
Actual colour of the blood in that vein
Dark red, maroon
at ~70% O₂ saturation. Computed live from its absorption spectrum. Never blue.
School diagrams print veins blue and arteries red, and everyone grows up "knowing" that deoxygenated blood is blue and turns red the instant it hits air. It doesn't. Draw venous blood into a syringe and it is dark red, a deep burgundy, the moment it leaves you. The blue you see through your skin is not the blood's colour. It's what a few millimetres of living skin does to light on the way in and out.
Operate the skin
Below is a patch of skin with a single vein running under it. The colours are not painted; they are computed, per wavelength, from real haemoglobin and melanin absorption data, through a light-transport model of the skin. Drag the sliders. Watch the surface hue drift, while the actual blood colour, pinned above, stays stubbornly red.
Light returning from OVER the vein
R … · G … · B …
more red than blue actually comes back
Dimming vs. the skin beside it
red −… · blue −…
red is subtracted; blue is barely touched → reads blue
Push it shallow and the blue drains out; red starts reaching the blood again, so the vein stops being red-deficient.
A thin vessel barely dims anything; it just reads as skin. Kienle's own vessels ranged turquoise → blue with size and depth.
Melanin darkens everything and hides faint veins; it changes how visible the effect is, not the direction of it.
Venous blood sits near ~70–75% (approximate; it varies by tissue); arterial is ~95–100%. Bright scarlet or dark maroon, never blue.
Two myths, not one
Myth one is the child's version: blood is blue until it meets air. Slain by the syringe, and by the pin above; the blood is dark red at every setting of that oxygenation slider. Human blood carries iron-based haemoglobin, which is red oxygenated and darker red deoxygenated, and nothing in between is blue.
Myth two is the grown-up "correction," the one you'll find on nearly every explainer page (and in most textbooks): blue light scatters back off the skin the way the sky scatters sunlight, so more blue reaches your eye; that's why the vein looks blue. It sounds authoritative. It is also wrong. In 1996 Kienle and colleagues put skin under a CCD camera and a Monte-Carlo simulation and measured what actually comes back. Their finding, which this model reproduces live: over a typical vein, MORE red light returns to your eye than blue. Look at the left readout: red beats blue at almost every setting.
So why blue? Because vision reports contrast, not absolute brightness. Red light dives deep, reaches the blood, and is soaked up, so the patch of skin over the vein sends back less red than the skin on either side of it. Blue light never gets that deep, so it returns from over the vein about the same as everywhere else. The vein is a place where red has gone missing. Against the redder skin around it, that red-deficit reads as blue. It is an illusion of subtraction and surround, and its sign flips with depth and diameter, which is why you can drag those sliders and drive the hue all the way back to red.
The check: every colour on this page, recomputed
Three sample wavelengths stand in for red / green / blue (650 / 550 / 450 nm). At each, the model needs three real numbers: how strongly blood absorbs, how strongly skin absorbs, and how strongly skin scatters. Nothing here is a stored colour; it is all the arithmetic below, run on every slider move.
| λ | ε HbO₂ | ε Hb | μₐ blood* | μₛ′ skin | μₐ skin |
|---|
*μₐ blood at the current oxygenation, cm⁻¹. Extinction ε in cm⁻¹/M (Prahl/OMLC). μₐ blood = 2.303·ε·(150 g·L⁻¹ / 64500 g·mol⁻¹).
For your current vein, the light budget per channel: reflectance over the vein vs. the skin beside it, and the dimming:
The two claims that matter, evaluated on your current settings:
The offline gate recomputes all of this two independent ways and asserts each claim: node research/why-veins-look-blue/verify-why-veins-look-blue.mjs.
What's exactly true, what's modelled, and every free choice
Exactly / physiologically true. Human blood is red at every oxygenation: bright scarlet arterial (~95–100% O₂), dark maroon venous. It is never blue. Venous ("mixed venous", SvO₂) saturation is roughly 70–75%, but that is a typical figure and genuinely varies by tissue and demand; treat it as approximate. The haemoglobin extinction spectra, the fact that red penetrates skin far deeper than blue, and the result that more red than blue returns over a vein, are all real, sourced measurements.
The scoped claim. "Never blue" is a statement about human blood. Some animals really do have blue or blue-green blood (horseshoe crabs, octopuses, squid, many crustaceans) because they carry copper-based haemocyanin instead of iron-based haemoglobin, and oxygenated haemocyanin is genuinely blue. So the precise claim is: human blood is never blue. Iron makes red; copper can make blue.
Modelled (a reduced-order stand-in for Kienle's Monte Carlo). The skin here is a Kubelka–Munk scattering background (dermis) under a thin Beer–Lambert melanin filter (epidermis), with the vein as a depth-weighted absorbing layer: the fraction of returning light that reaches vein depth d is exp(−2·μ_eff·d), and that fraction is attenuated by passing a chord ≈ the vein diameter, exp(−μₐ_blood·D). This is not the full Monte-Carlo photon transport of the paper; it is the smallest honest model that reproduces the paper's two qualitative results: (1) absolute red reflectance over the vein exceeds blue, and (2) the vein reads blue by red-deficit contrast, with the hue flipping as depth/diameter shrink. The pinned actual blood colour is computed directly from the blood's own absorption spectrum and is the quantitatively honest part.
Free choices & constants. ε(HbO₂), ε(Hb) at 450/550/650 nm from the Prahl/OMLC compilation; whole-blood haemoglobin 150 g·L⁻¹, MW 64500. Skin reduced scattering μₛ′(λ)=46·(λ/500)−1.421 cm⁻¹ and baseline dermal absorption 7.84×10⁸·λ−3.255 cm⁻¹ from Jacques 2013; melanin absorption 6.6×10¹¹·λ−3.33 cm⁻¹ (Jacques 1998) filtered through a 0.1 mm epidermis; a fixed 0.5% background dermal blood at 75% O₂; blood scattering μₛ′≈25 cm⁻¹ for the pinned swatch; round-trip depth factor 2. Colours are a three-sample (R/G/B) proxy rendered through the sRGB transfer curve, not a full CIE integral; enough to show the drift honestly. The perceived-hue label is derived from the Retinex-style vein/skin reflectance ratio, the same contrast quantity Kienle invokes.
Not medical advice. This is a page about optics, not veins-as-symptoms. It says nothing about visible, prominent, or varicose veins as a health matter; for anything like that, ask a clinician, not a light-transport toy.