In the first round at Wimbledon in 2010, John Isner hit 113 aces in a single match. Mahut hit 103. The match ran 11 hours and 5 minutes across three days and finished 70-68 in the fifth set. Those numbers get quoted constantly, usually to make a point about serving. They make a different point about tennis terminology: a stat is only as clean as the rule that defines it, and "ace" is messier than it looks.
This piece is about one term and one number. We start with what an ace is, then look at what that 113 actually counted, then at what it quietly leaves out. By the end the goal is not a serving tip. It is the ability to read a stat line without overtrusting it.
What counts as an ace in tennis
An ace is a legal serve that the receiver fails to touch with the racquet, ending the point immediately in the server's favor. Three conditions all have to hold. The serve must land in the correct service box. The receiver must not make contact with the ball. And it must be the serve that ends the point — a serve that clips the net cord and still lands in but goes untouched is an ace; a let is replayed and counts as nothing.
That third condition is where casual fans slip. If the returner gets a frame on the ball and shanks it long, it is not an ace. It is a service winner, or in most scoring systems simply an unreturned serve. The defining feature of an ace is the absence of contact, not the speed of the serve. A 90 mph kick serve that handcuffs a returner who never moves the racquet is an ace. A 140 mph flat serve the returner blocks into the net is not.
What the 113 actually measured
Here is the part the headline number hides. Isner's 113 aces did not happen because he served 113 unreturnable balls in a normal match. They happened because the fifth set had no tiebreak and went to 70-68. The match contained roughly 980 points. The longer a match runs, the more aces accumulate, because aces are a count, not a rate.
Walk it through in order, the way the point actually unfolds.
The mechanism, step by step
First the toss and the contact height. Isner is 6 feet 10 inches. Contact on a flat first serve happens somewhere above 11 feet, which steepens the angle into the box and shortens the time the ball spends in the air on the returner's side. Then the ball speed: a first serve in the 130–140 mph range covers the roughly 60 feet from baseline to baseline in well under half a second. Then the return window. By the time the ball crosses the service line, the receiver has had a few hundred milliseconds to read direction, decide, and move the racquet into the contact zone. On grass, which is fast and keeps the bounce low, that window narrows further.
None of that explains 113 on its own. A tall man serving fast on grass against a good returner produces a high ace rate. The format produced the total. Strip out the marathon fifth set and the number stops being a record and becomes a very good day. The terminology — "most aces in a match" — quietly bundles a skill (serving) with a circumstance (an untiebreaked set that ran past 130 games). The stat reports the sum and stays silent about the split.
What the number doesn't measure
An ace count is not a clean measure of serving quality, and treating it as one is the common mistake. Four things confound it, and they all push in different directions.
Surface. Grass and fast hardcourts inflate ace counts; clay suppresses them by slowing the ball and lifting the bounce into the strike zone. Comparing a player's Wimbledon aces to their Roland-Garros aces tells you more about the courts than the player.
Format and opponent. A best-of-five against a passive returner gives more serves to ace than a best-of-three against a returner who stands deep and blocks everything back. More serves, more chances. The count rises without the serve improving.
Scoring context. As the Isner match showed, the no-tiebreak final set was the entire reason the total reached three figures. Since 2022 all four majors use a final-set tiebreak at 6-6, which caps how long sets can run. The 113 cannot be broken under current rules in the same way. That is a terminology problem disguised as a records problem: the thing being counted changed when the rulebook did.
What an ace replaces. A serve-plus-one point won in two shots looks "worse" in the box score than an ace, but it can reflect the same dominant serve drawing a weak return. The ace gets the glory line; the service winner does the quiet work. Counting only aces undervalues serves that are good enough to win the point but not good enough to go untouched.
So the honest reading of 113 is: an extraordinary serving performance, on the most ace-friendly surface, in a format that no longer exists, against an opponent who himself hit 103. All of that is true at once. The number is real. Its meaning is conditional.
Aces among their neighbors
Most stat-line confusion comes from terms that sit next to "ace" and get used interchangeably when they shouldn't be. Here is the practical distinction.
| Term | Receiver touches ball? | Point outcome | Counts as |
|---|---|---|---|
| Ace | No | Server wins instantly | Ace (and a service winner) |
| Service winner | Sometimes | Server wins in one or two shots off the serve | Service winner, not always an ace |
| Unreturned serve | No (or shanked) | Serve not put back in play | Broad category that includes aces |
| Let | No (net cord, lands in) | Point replayed | Nothing |
| Double fault | No | Receiver wins point | Server error, not a return stat |
Every ace is a service winner. Not every service winner is an ace. And the double fault — two consecutive faults losing the point outright — is the ace's mirror image: it sits in the same serving box score but moves the point the other way. A player with a gaudy ace count and a quietly large double-fault count is not the server the ace number alone suggests.
A rule of thumb for reading your own ace count
If you track your own matches, read aces as a rate, not a trophy. Count aces per service game, or per first serve made, before you compare anything to anyone. Twelve aces in a three-set marathon on a fast court is a different serve than twelve aces in a tight 6-4, 6-4 on clay, and the per-game rate will tell you which.
Then compare like with like. Your indoor-hardcourt number does not belong next to a friend's clay-court number. And before you read an ace count as a sign your serve improved, check whether you simply served more games, or played a returner who stood farther back. The count moves for reasons that have nothing to do with the toss.
One more, stated plainly: a falling ace count is not automatically a worse serve. If your double faults dropped at the same time, you may have traded a few risky flat first serves for a higher first-serve percentage. The box score lost some aces and the match got easier to win. That is a good trade the ace line will never show you. A note from one of our reviewers. In my own practice I stopped charting aces and started charting first-serve points won, because the day I served best all year — flat, into the body, returner late on everything — produced exactly two aces and a 78 percent hold rate. The serves that won me the match were the ones the count refused to record. That is the whole argument of this piece, lived out in one stat line: the number you can see is rarely the number that decided the point.