Replace one dead bulb in a fixture of three and you have started an experiment. The new tube of glass promises the same "warm white, 2700 K" as the survivors, and yet side by side one glows a touch pinker, or greener, or cooler, and now the mirror over the sink is lit like a museum of slightly different mornings. Neither bulb is defective. The color on the box is not a single number that a factory must hit exactly; it is the center of a tolerance zone, and within that zone—plus a second, less-advertised axis called tint—the two bulbs are legally, standardly, minding-their-own-business different. Matching bulbs is not a matter of reading the same label twice, which is the quiet truth inside the way warm light reshapes a whole room: the light you buy is a wider family than the word on the carton admits.
A name that is a bin, not a point
Color temperature, for lights, is defined by chromaticity coordinates—the technical name for where a light's color lands on a two-dimensional map. The American National Standard for solid-state lighting chromaticity, ANSI C78.377, draws a quadrilateral region around each nominal value: every bulb whose measured color falls inside the 2700 K quadrangle is a 2700 K bulb. Those regions are deliberately roomy, because manufacturing phosphor-coated LEDs to tighter tolerance would cost real money, and for a single bulb standing alone the width of the zone is invisible. Put two bulbs side by side in one fixture, though, and you have built a comparison instrument: human vision is far better at judging two lights together than either light alone, precisely the effect behind street lamps changing color over an evening walk—side-by-side comparison turns tiny differences into obvious ones.
Two doors out of the problem exist, and both cost a little intention. The first is shopping by tint specification rather than temperature alone: listings aimed at specifiers publish the tint deviation explicitly, and matching that number between bulbs buys you the same side of the warm-white line. The second is buying matched sets and installing them together, so the fixture's population never mixes eras. Neither is advertised on the front of the box, because the box is printed for aisles, not for bathrooms—and the aisle, lit by its own sea of mixed lamps, is the one place on Earth you cannot judge bulb color at all.
Two directions away from perfect
The second axis is the sneaky one. A warm white can sit exactly on target in the orange-to-yellow direction and still drift in the perpendicular direction—above the line of "perfect" warm whites toward green, or below it toward pink. The industry measures this drift as Duv, and energy-certification programs limit how far a qualified bulb may wander, but within the limit both directions are allowed. That is the shape of most bathroom mismatches: same 2700 K on the box, one bulb a half-step green, one a half-step rose. Aging adds a slow third act, since LEDs drift slightly in color over thousands of hours, so the survivors of an old set have moved since installation while the new bulb arrives fresh from the center of its zone.
Rendering rides on top of tint
Two matched-tint bulbs can still disagree on your face. Color rendering—how completely a light's spectrum feeds the reflectance of skin, wood, and towels—varies by phosphor recipe, and a bulb that renders reds richly will make the same beige wall feel warmer than its cheaper shelf-mate of identical color temperature. Render differences are invisible in the aisle and obvious at home, which is why spec sheets list a rendering index and why lighting programs like the Department of Energy's purchasing guidance treat color quality as a first-class criterion, not a footnote. Brightness completes the illusion: a slightly dimmer bulb usually reads slightly warmer, so even two identical bulbs on different dimmer settings will refuse to match.
Buying bulbs that agree
The practical playbook is short. Replace lamps in one fixture as a set, not one at a time—manufacturers aim at the same zone but drift between production runs. Buy the same brand and model line, and when the carton offers it, match the tint specification as well as the color temperature. Install and judge at home in the evening, in the actual fixture, because store lighting makes honesty impossible. And when one bulb of a beloved mismatched pair finally dies, accept the physics: you are not shopping for a number, you are shopping for a neighbor. A quick eye for two nearly identical lights is trainable, and it is the exact skill exercised in Speed Color Match, where telling two close colors apart, quickly, is the entire game.
A diagonal line marks perfect warm whites; the tolerance box around 2700 K contains both dot A below the line (rose cast) and dot B above it (green cast). Labels tell the story even without hue.