1. Sunlight Takes About 8 Minutes and 19 Seconds to Reach Earth
Two values have exact definitions agreed on internationally. The speed of light in a vacuum is 299,792,458 meters per second [1]. The astronomical unit, a ruler for the Solar System that was first meant to be the average distance from Earth to the Sun, is 149,597,870,700 meters [5]. NASA gives the same distance as roughly 93 million miles (150 million kilometers) [4].
Divide the distance by the speed and you get about 499 seconds, which is 8 minutes and 19 seconds. That division is our own arithmetic from the two sources, not a number printed in them. You will also see "about 8 minutes 20 seconds," which is the same thing rounded.
One catch: Earth's orbit is a slightly squashed circle, so the real distance changes through the year. The 8 minutes 19 seconds is the trip over the average distance. The next question is the interesting one: for most of history nobody could even show that light takes any time at all.
2. On Earth, Light Crosses Any Distance Too Fast to Time by Eye
Try timing a flashlight beam across a field. Across 10 kilometers (about 6 miles), light needs 10,000 ÷ 299,792,458, or about 0.00003 seconds. That is roughly one thirty-thousandth of a second. This is our own calculation. A person with a stopwatch could not time it accurately, so the delay looked like zero.
To see the delay, you need a trip so long that the delay grows large. That is the trick the first successful measurement used: look for a distance big enough that light would obviously be late. The Solar System has plenty.
3. Jupiter's Moon Io Showed Up Late When Earth Was Farther Away
In the 1670s, astronomers in Paris, including the Danish astronomer Ole Rømer, were timing the eclipses of Io, the innermost of Jupiter's four big moons. Io swings around Jupiter in about 42.5 hours and slips into Jupiter's shadow on every lap, so it works like a clock in the sky. Rømer and his colleagues kept watching these eclipses between 1671 and 1673 [3].
The eclipses seemed to drift from the predicted times. When Earth was moving away from Jupiter, each eclipse was seen a little later than expected. When Earth was moving closer, it came a little sooner. Rømer's explanation was that the light from Io has to travel farther to reach us when Earth is farther from Jupiter, so the eclipse is seen later [3].
He then put it to a test. In September 1676 he predicted that an eclipse due on November 9 would arrive 10 minutes late compared with the standard tables. The eclipse was then seen about that much late, as he had predicted [3]. Later that month he told the Royal Academy of Sciences in Paris that light needs about 22 minutes to cross the full width of Earth's orbit [3].
That is 11 minutes for half the width, which is about the distance from the Sun to Earth. Next to today's 8 minutes 19 seconds (499 seconds), 11 minutes (660 seconds) is about a third too long. Given 1670s clocks and an uncertain size for Earth's orbit, that is a remarkable first estimate.
A detail that surprises many readers: according to a history article from Optica, Rømer never wrote down a speed for light. What he gave was the 22 minutes [3]. Turning that into a speed takes a size for Earth's orbit, and with the modern size, 22 minutes works out to about 227,000 kilometers per second, roughly a quarter below the true value [7]. Not everyone accepted the idea right away, either. Cassini, the director of the Paris Observatory, was never fully convinced [7].
4. In 1849, a Spinning Toothed Wheel Timed Light on Earth
It took until the middle of the 1800s to time light without leaving the ground. In 1849 the French physicist Hippolyte Fizeau set up a mirror on Montmartre in Paris and sent a beam toward it from Suresnes, about 8,633 meters away. In the path of the beam he put a rapidly spinning toothed wheel [2].
A simple way to picture it: light slips out through a gap between two teeth, bounces off the far mirror, and comes back. If the wheel spins just fast enough, a tooth has moved into the gap by the time the light returns, and the beam is blocked. So the spinning speed at which the beam vanishes tells you how long the light took for the round trip [2].
Fizeau's result was off by about 5 percent, which the biography at the University of St Andrews calls almost as accurate as Rømer's astronomical measurement [2]. A measurement made with a wheel in a city and a measurement made with a moon of Jupiter gave answers in the same neighborhood. Two such different methods agreeing is what makes scientists trust a number.
5. Since 1983, the Speed of Light Is Fixed and the Meter Comes From It
Measurements kept getting better. In 1983, the international body that sets measurement standards decided to stop measuring the speed of light and instead fix it. Its resolution defined the meter as the length of the path traveled by light in a vacuum during 1/299,792,458 of a second [6]. The value 299,792,458 had been recommended in 1975, and the new definition kept it unchanged, so the length of a meter did not jump [6].
The wording was refreshed in the 2018 revision of the SI, but the speed of light is still exactly 299,792,458 meters per second [1]. So the speed of light stopped being something to find out and became the ruler itself. This is why the 8 minutes 19 seconds above is calculated from the defined distance and speed, rather than measured directly. If you are curious why scientists gave up the old metal bar, our article Why Did the Definition of the Meter Change from a Metal Bar to Light? follows that story.
6. The 8 Minutes Only Counts the Trip After Light Leaves the Sun
One more point keeps the story honest. The 8 minutes 19 seconds is the travel time from the Sun to Earth through nearly empty space. It says nothing about how long the energy took to get out of the Sun first. Inside the Sun, the gas is so crowded that light keeps getting absorbed and sent off in new directions, so energy made deep in the core takes far longer to reach the surface than the open-space trip does. Our article Why Does Light From the Center of the Sun Take Tens of Thousands of Years or More to Reach Its Surface? goes into that part.
7. Try It Yourself: Do the Division and Watch Jupiter's Moons
Grab a calculator and divide 149,597,870,700 by 299,792,458. If it shows about 499, you have calculated the travel time from the Sun to Earth yourself. For a smaller version, divide 1 by 299,792,458. The answer is about three billionths of a second, the time light takes to cross one meter.
If you have binoculars and a clear night when Jupiter is up, you may see its four big moons as little dots in a row, and they change position from night to night. Rest your arms on something steady. Never point binoculars or a telescope at the Sun. Watching the moons shuffle is a small taste of the clock Rømer used.
If you want to go deeper, the Optica history article on Rømer is a good place to see how he tied together nine years of eclipse timings [3].