What colour lure can the fish you are after actually see?

Red light is down to 1% of its surface strength somewhere between 1.2 m and 11 m, depending entirely on the water: 11 m in the clearest ocean, 1.2 m in flood water. But absolute range is the less useful half of the answer. What matters is which colour survives best, and that flips: the wavelength that penetrates furthest is about 470 nm — blue-green — in clear ocean, and about 680 nm — deep red — in a tannin-stained lake, because dissolved organic matter absorbs short wavelengths hardest. In tannic water red is the colour that wins, which is the opposite of the usual advice.

It matters for whether a fish can detect the lure at all, which is a narrower claim than most advice makes. This site computes the contrast between a lure and the water behind it, and the distance at which that contrast rises above the species’ detection threshold. It cannot tell you whether a fish will strike: that is behaviour, and no optical model predicts it.

It depends on which way the fish is looking, which is why the popular advice contradicts itself. Looking sideways or down, brightness wins and chartreuse or white stand out. Looking up against the surface, nothing can be brighter than the sky, so only a dark silhouette makes contrast — which is why black and blue also "always produce" in muddy water. Both camps are right about different geometries.

Most of them see two colours where you see three. Of the 54 species here, 19 have direct spectral measurements of their cones; trout, salmon and carp have four cone types including one sensitive to ultraviolet, and sharks have one and are colour-blind. Each of those sorts the same tackle box differently.

About 470 nm in clear ocean water and about 680 nm in a tannin-stained lake — blue-green in one, deep red in the other, computed from the water’s own absorption rather than read off a chart. There is no single answer, and a chart that gives one is describing a water it did not name.