Has the Speed of Light Changed Over Time?

The short answer depends on what you mean by “the speed of light changing.” The number written in meters per second is fixed by the modern definition of the meter. The deeper physical question is whether light’s behavior, compared with other physical processes, was different in the distant universe. That question can be tested, but there is no confirmed evidence that it was.

Why the number is exact today

The vacuum speed of light is denoted by $c$. In the International System of Units (SI), its value is defined to be exactly

$$c = 299\,792\,458\ \text{m/s}.$$

Since 1983, the meter has been defined using this exact value and the SI definition of the second. In simplified terms, one meter is the distance light travels in vacuum during $1/299\,792\,458$ of a second. So modern experiments do not measure the SI value of $c$ with uncertainty; they use it to define the meter and measure other quantities against it.

Before that change, scientists measured light’s speed and gradually improved the precision. The reported number shifted slightly as measurement techniques improved. Those shifts were changes in the estimate, not evidence that light itself had sped up or slowed down.

Why a changing number is not enough

Meters and seconds are human-defined units. If the length of a meter or the duration of a second were defined differently, the numerical value of $c$ would also be different—even if no physical process had changed. For this reason, a claim that a dimensional quantity such as $c$ changed needs to specify how it is compared with other quantities and how the units are defined.

A useful test compares dimensionless constants, which have no units and therefore do not depend on choices such as meters versus feet. One example is the fine-structure constant, usually written as $\alpha$. It helps set the strength of electromagnetic interactions. In SI notation,

$$\alpha = \frac{e^2}{4\pi\varepsilon_0\hbar c},$$

where $e$ is the elementary electric charge, $\varepsilon_0$ is the vacuum permittivity, and $\hbar$ is the reduced Planck constant. This formula illustrates why a proposed change in $c$ cannot be interpreted in isolation: the physically testable claim concerns how dimensionless combinations of constants behave.

How astronomers look for changes

Light from distant galaxies and quasars carries spectral lines produced by atoms and molecules. The pattern and spacing of those lines depend on physical constants. Researchers compare observed spectra with laboratory measurements to test whether dimensionless quantities, including $\alpha$ or the proton-to-electron mass ratio, differed when the light was emitted.

For instance, suppose a distant source shows a spectral pattern that appears slightly different from the laboratory pattern. That alone does not establish that $c$ changed. Researchers must account for the source’s motion, the expansion of the universe, the intervening material, instrument calibration, and the physical model linking the spectral pattern to constants. Different measurements and independent checks are needed to distinguish a real variation from these effects.

These observations place limits on possible changes over large spans of cosmic time. Some measurements have prompted debate about tiny variations, but there is no broadly accepted, confirmed detection. A bound on a dimensionless constant is also not automatically a bound on $c$ alone; interpreting it requires a model that says how the constants are related.

What about the early universe?

Some speculative cosmological theories propose that the speed of light, or the laws governing it, differed in the early universe. Such proposals are sometimes called varying-speed-of-light models. They are alternatives or extensions to standard cosmology, not an established result that light’s speed has been observed to change.

There is also a separate issue in general relativity: the coordinate speed assigned to light can depend on the coordinate system used to describe curved spacetime. Light can also travel more slowly through a material than through vacuum. Neither effect means that a local observer measuring light in vacuum finds a different value for $c$; in relativity, that local measurement is $c$.

A concrete way to interpret the claim

Consider two statements:

1. “The speed of light is 299,792,458 meters per second.” This is exactly true by the current SI definition.
2. “The fine-structure constant had a different value billions of years ago.” This is a physical hypothesis that can be tested by comparing astronomical spectra with laboratory data.

The first statement is about units. The second is about whether physical relationships have changed. If a study reports evidence relevant to a changing $c$, check which dimensionless quantity it actually measured and what assumptions connect that result to $c$.

Common mistakes

• Treating the exact SI value as proof that nature never changes. The exact value follows from the unit definition; it is not a measurement of the universe at every time and place.
• Taking old and new tabulated values as evidence of cosmic variation. Improvements in experimental methods can change a measured estimate without any change in the physical quantity.
• Equating light’s speed in matter with its vacuum speed. Light propagates at different speeds in materials, but that is not evidence that the vacuum value changed.
• Saying “c changed” without specifying an observable. The meaningful empirical question is whether dimensionless physical relationships changed and whether measurements support that claim.

The best current answer is therefore: the numerical SI value is fixed by definition, and no change in the underlying physical behavior of light over cosmic time has been confirmed. Cosmological variation remains a testable, model-dependent possibility, not an established fact.