The serve speed displayed on the scoreboard after every point is one of those numbers fans take at face value. 137 mph. 142 mph. The display updates instantly, so it must be measuring something precise. It is — but what it is measuring varies slightly by venue and by system, which is why the same player's serve can read 5 mph different at two tournaments in the same week. Here is how it actually works.
Where the speed is measured
Tennis serve speed is measured at the point of impact — the moment the racket strikes the ball — not at any later point in the ball's flight. This matters because the ball decelerates significantly after leaving the racket. By the time a serve reaches the returner, it has lost roughly 30–40% of its speed.
A 130 mph serve at impact is, by the time it bounces, roughly 90 mph. By the time the returner reaches it, it is something like 70–80 mph depending on the surface. This is part of why fast surfaces (grass) feel so different from slow surfaces (clay) at the receiving end. The radar number is the same, but the ball that reaches the returner is not.
The technology: radar guns and Hawk-Eye
For decades, tennis serve speed was measured by a radar gun pointed at the server from behind the baseline. The gun emits a microwave signal, the signal reflects off the ball, and the Doppler shift in the returning signal is converted into a speed reading. This system is simple, fast, and accurate to within roughly 1 mph when calibrated correctly.
Modern tournaments increasingly use camera-based ball tracking systems — Hawk-Eye being the dominant one — which calculate ball trajectory from multiple high-speed cameras. These systems can compute speed at any point in the ball's flight, not just at impact, which is why some broadcasters now show both "serve speed" and "speed at the baseline" as separate stats.
Why the numbers vary between venues
You may have noticed that a player's typical first-serve speed varies a bit from tournament to tournament. There are several reasons.
- Radar placement and calibration: radar guns are calibrated to within a tolerance, but the tolerance is not zero. Two different guns at two different venues can show a 2–3 mph difference on the same physical serve.
- Altitude: serves at high-altitude tournaments (Madrid, for example) travel faster because the air is less dense. The same physical effort produces a higher reading.
- Ball condition: newer balls fly faster. By the seventh game of a set, balls have been hit dozens of times and have lost some of their compression.
- Court speed: this affects the perceived speed and the speed at impact with the returner, but not the radar reading at impact, which is purely about racket speed.
The official records
The ATP currently lists Sam Groth's 263 km/h (163.4 mph) serve at the 2012 Busan Challenger as the fastest officially measured serve in men's professional tennis. The asterisk is that the radar gun in Busan was not the standardized ATP-tour gun — different equipment, different calibration. The fastest serve at an ATP main tour event is John Isner's 253 km/h (157.2 mph) at the 2016 Davis Cup, with various other recorded serves in the 150+ mph range on tour.
On the women's tour, the official record is Georgina Garcia Perez's 220 km/h (136.7 mph) at a 2014 Korea Open match, though Sabine Lisicki had a 211 km/h (131 mph) serve at Stanford in 2014 that is sometimes cited. The women's record numbers are not tracked as obsessively as the men's, partly because the gap between the top men's server and the top women's server is consistent and well-understood.
Why speed isn't the whole story
Two serves at 130 mph are not the same serve. Placement, spin, and the angle of delivery matter at least as much as raw speed. A 120 mph serve out wide to the deuce-court forehand of a right-hander is much harder to return than a 135 mph serve straight at the body. This is why the top servers in the world — players like John Isner, Reilly Opelka, or in earlier eras Pete Sampras — are not always the players with the fastest measured serves. They are usually the players who combine high speed with strong placement and disguise.
The serve-speed stat is fun and easy to display, but the more useful numbers are first-serve percentage and percentage of first-serve points won. Those tell you how effective a serve is, not just how fast.