The blue in a newborn's eyes is a first draft. Melanin—the same brown-black pigment that tans skin and colors hair—has barely been deposited in the iris at birth, and most babies start life with some version of slate blue or gray. Over the following months, cells in the iris ramp up pigment production under genetic instructions, and the draft gets revised: darker for most babies of African, East Asian, and Indigenous descent, sometimes holding pale for babies of European descent, with green and hazel occupying the middle lanes. Pediatric ophthalmology guidance puts the usual settling point around six to twelve months, with some fine-tuning into the toddler years. Whether a particular pair of eyes ends up brown or blue depends on how much melanin that schedule delivers—and on a piece of physics that surprises most parents: no human iris contains blue pigment at all. Blue eyes are blue the way the ocean is blue—by what they do with light, not by what pigment they hold.
The iris is a layered light filter
Anatomically, the iris is a small stack of layers doing two jobs: controlling light and coloring it. The front border layer is packed with melanocytes, cells that load melanin into granules. Behind them lies a stroma of collagen fibers, and behind that a pigmented back layer that nearly everyone has, regardless of eye color. Color emerges from how light negotiates this stack. Melanin absorbs light broadly; the collagen scatters shorter, bluer wavelengths more strongly—the same preferential scattering that colors the sky. With little melanin up front, incoming light scatters blue from the stroma and much of it bounces back out: blue eyes. With dense frontal melanin, the brown granules absorb and brown the returning light before blue scatter can organize: brown eyes. Hazel and green report mixed loads and layer thicknesses. The back layer, always dark, deepens the effect by absorbing light that would otherwise wash the scattering out.
Melanin arrives on a schedule
The reason newborns start pale is developmental. Melanocytes exist in the newborn iris, but they are understaffed with pigment—a state called hypopigmented. Over the first months, enzyme activity rises and granules accumulate, so a six-month-old iris may hold several times its birth melanin. The visible trajectory is therefore one-way: baby blues darken or hold, and eyes essentially never lighten dramatically later, because pigment that has been deposited does not withdraw. This schedule is why the classic advice to "wait a year" before assigning the nursery's eye-color verdict is sound, and why preemies, who started the clock earlier, can keep changing a bit longer after their due date.
Genes hand the recipe, not the color
School genetics taught a tidy story—brown dominant, blue recessive—that genome research has thoroughly complicated. Two genes near each other on chromosome 15, OCA2 and HERC2, dominate the outcome, but dozens of additional variants shift the amount and distribution of melanin, which is why two brown-eyed parents can produce a blue-eyed child and why hazel and green exist at all. Genetic studies of iris color describe it as a polygenic, continuously varying trait—prediction from DNA is probabilistic, not deterministic. The baby's final color was negotiated in this genetic committee long before conception; birth just delays the announcement while the pigment ships.
Blue without any blue pigment
The no-blue-pigment point deserves a moment, because it reframes every eye color conversation. Brown and hazel eyes are colored by melanin, an actual pigment. Blue eyes—and much of the blue-gray of the newborn—are structural color: the geometry of transparent fibers selecting which wavelengths bounce back. Newborn slate blue reads grayer than adult sky-blue because infant stroma scatters from a slightly different collagen arrangement and melanin sits unevenly while arriving; as pigment lands, the same physics starts producing the clearer adult hues. It also explains lighting's power over pale eyes: against a blue shirt, in shadow, or under gray sky, a pale iris can shift a shade—the same eye, different illumination. Watch how borderline shades fight with language as you try to name them—teal versus slate versus gray—and you have landed on the boundary that Blue or Green? exists to explore.
What change is worth watching
Most color change is unremarkable biology. Two situations merit a professional look, and both are about difference, not color: eyes that are two clearly different colors (complete heterochromia) or one iris that develops a patch—usually harmless, occasionally a sign of pigment change worth checking—and any white or markedly asymmetric pupil reflection in flash photos, which pediatricians ask parents to report. Neither is about whether the baby went blue or brown; the timetable itself is broad and healthy across its whole range. A fast-darkening iris and a slow one are both normal, and the speed of the change says nothing about how well the eyes work.
The settled color, whatever it is, will owe nothing to a vial of blue dye. It will be a melanin count and a scattering geometry—a layered filter that spent its first year being written, and then spends a lifetime doing something quietly remarkable every time someone looks into it.
Each disc mixes a blue-gray scatter tone with increasing amounts of brown melanin. In some infants the third disc never darkens—the blue stays.