Arrow speed is the number that gets quoted most on the range, and anyone with a chronograph writes it down next to their draw weight. Behind it sits a simple question: how many metres a second does my arrow fly, and if I can make it go faster, will I score more? Studies answer the first half with fairly precise figures, taken with chronographs, microphones, radar and high-speed cameras. They have much less to say about the second half, because almost all the measurements that link speed to the point of impact were made with shooting machines, and nobody has ever checked, within the same archer, whether a faster arrow is worth points.
How many metres a second
For the Olympic recurve the firmest figure is a time rather than a speed. In thirty-nine elite archers from four countries, filmed with high-speed cameras, the arrow covered seventy metres in 1.24 seconds, with a standard deviation of six hundredths; in twenty archers from the Turkish national squad and its candidates, timed with a field chronometer, it took 1.29 seconds, and a week later each archer came back with almost the same value. Seventy metres in 1.24 seconds is an average of about fifty-six metres a second, but that average covers the whole flight: the arrow is faster as it leaves the bow, and a calculation for a 202 newton Olympic bow launches two different arrows at 59.5 and 62.6 metres a second. In the registry the claim that at seventy metres an elite archer's arrow arrives in a little over a second, at an average of fifty to sixty metres a second is reported, because there are two measurements and the first does not reconcile the time and the speed it gives.
Laboratory tests give the same order of magnitude. Recurve arrows leave the bow at roughly 57 to 65 metres a second, whether an archer or a machine is shooting, and this is the part that holds up in the claim that a recurve competition arrow weighs about fifteen grams, is some sixty centimetres long and leaves the bow at around sixty metres per second. The entry is contested over weight, not speed, because fifteen grams describes a shaft such as the A/C/E, while X10s weighed on the bench come in between 19.7 and 21.2 grams. I cover weight and spine in the article on choosing competition arrows.
The compound sits higher. In the laboratory four compound bows launch the arrow at between 70 and 82 metres a second, and the only comparison drawn from a competition sample, 61 archers at a 2025 World Cup stage, gives 292.0 against 216.7: the units are printed as metres per second, but they only make sense as feet per second, which is about 89 against 66 metres a second. That compound arrows leave the bow much faster than recurve arrows is reported: the direction is not in doubt, but nobody has measured how large the advantage is with the same draw weight and the same arrow.
Wooden bows sit at the other end of the scale. A 48 lb elm flatbow, held in a mechanical rig, sends wooden arrows off at between 42.66 and 43.73 metres a second. A replica yew bow drawn at around 16 kilos, with a 9.7 gram cane arrow, stays between 33.5 and 39.2 metres a second, and in the model calibrated on those shots the arrow loses half its energy within the first fifteen metres. On that single series of tests the registry reports that a medium-weight wooden bow with a very light cane arrow does not exceed forty metres per second and only works at close range.
All these figures also depend on how they were taken. Optical chronographs, microphones, radar, cameras and sensors inside the shaft do not all measure the same thing, and the only comparison against a metrology reference found three commercial chronographs reading between 1.5% and 1.9% low, all in the same direction and to the nearest whole metre per second. This is why the claim that commercial optical chronographs read arrow speed with an error of a couple of per cent is reported, from four shots with a compound. How each instrument works is explained on the page about how arrow speed is measured.
What decides launch speed
The speed at which the arrow leaves the string depends on how much energy the bow stores and on how much of that energy reaches the arrow. The first is set mainly by draw weight. On a laminated recurve mounted on a machine, the same 23 gram arrow goes from 45.5 metres a second at 40 lb to about 54.3 at 50 lb, measured twenty metres out. Across different archers there is a single study, the one on the 61 World Cup archers: declared draw weight correlates with speed at 0.854, but recurve and compound are analysed together, and bow type on its own correlates with speed at 0.787. That among archers, those who shoot heavier draw weights send the arrow off faster is therefore reported: the mechanics are clear, but nobody has changed the draw weight of the same archer to see how much speed goes up.
The second is largely down to the arrow. With the same bow a heavier arrow leaves more slowly but takes away more energy, because part of the stored energy stays behind moving the limbs and the string, and that share shrinks as the arrow gets heavier. On the same recurve at 50 lb, going from 23 to 33 grams costs the arrow about 3 metres a second, but its energy rises from 24.4 to 32.4 foot-pounds; on a compound with a 20.6 gram arrow, 80% of the energy the bow gives back reaches the shaft. That with the same bow a heavier arrow receives more energy remains reported because nobody has measured the curve on an Olympic recurve with competition arrows, over the weight ranges used in competition. Between two shafts a quarter of a gram apart, shot by the same archer, the heavier one leaves at 59.21 metres a second against 59.67 and the energy is practically the same.
The rest of the set-up moves things less. The string matters through its mass, since about a third of it is added to the arrow's mass when speed is calculated. On material, the two tests available point the same way: in kyudo synthetic fibres pass about 69% of the energy to the arrow against 67.7% for hemp, and on a recurve held in a rack the same arrow leaves at 49.73 metres a second with Dacron and at 52.09 with Fast Flight. On the number of strands, though, one test finds a gain of 0.3 to 1.1 feet per second out of about 215 and the other no difference at all, which is why the claim that the string's material and number of strands change the arrow's speed is contested. Vanes change the launch very little: on a shooting machine there is about 1.5% between the bare shaft and the slowest vane, and with the wooden bow five different fletchings stay within just over a metre a second, although the differences grow later in flight. That vane type and size change the arrow's launch speed only slightly is reported, from two tests without statistical analysis.
How much speed it loses on the way
Once it has left the bow the arrow slows steadily because of air resistance, and this is the only claim on this page that the registry holds as corroborated: along its flight the arrow loses an appreciable part of its launch speed. An electronic point on a compound arrow records a drop from about 82 to 78.2 metres a second over fourteen metres; an inertial sensor inside the shaft measures about 8% lost over 55 metres indoors, and the same authors note that over a seventy-metre range the loss is around 15%. How much an arrow slows depends on the shaft, the point and the fletching, and it is the reason the average of about fifty-six metres a second over seventy metres sits below the launch speed. Where the drag comes from is covered in the article on how much the air slows an arrow.
How far one metre a second moves the arrow on the target
With the same aim, a faster arrow arrives sooner and drops less, so it strikes higher. The practical point is by how much. On a compound launching the arrow at 74.5 metres a second, measured with a machine in the laboratory and on the range and checked against a ballistic calculation, one extra metre a second raises the point of impact at seventy metres by about thirteen centimetres: almost twice the effect of a one-metre error in distance or of 20% more aerodynamic drag, each worth about seven centimetres. With a slower arrow the effect grows. In a model for aiming a traditional bow, with the arrow at 60 metres a second, three metres a second more raises the impact at seventy metres by 80 centimetres and three less lowers it by 69, which is more than twenty centimetres per metre a second.
The measurements confirm the direction, with a caution on the size. In high-speed footage of launches over fifty-five metres, the vertical deviation at the target matched the vertical component of launch speed multiplied by the flight time: the agreement was good with a robot and weaker with the only archer filmed, where variation in horizontal speed also played a part. Across fifty-nine X10 arrows shot by a machine with a 52 lb recurve at seventy metres, mass and mean spine accounted for 46.7% of the vertical spread, with the heavier arrows landing lower, as you would expect given that a heavier arrow leaves the same bow more slowly. For the 0.282 grams between the lightest and the heaviest, however, the ballistic calculation predicted more than fourteen centimetres of vertical range, and the measurement gave 8.8.
This is why the claim that small differences in launch speed account for much of the vertical spread is reported: the direction is firm, the calculated size needs a margin, and the link has never been measured in a group of archers. The same machine test supports another entry, that from the same bow, a heavier arrow strikes lower, which is also reported.
How much speed changes from one arrow to the next
What affects the group is speed changing from shot to shot, more than the average speed, and in expert archers it changes very little. In seven archers from the German national squads, over thirty arrows at eighteen metres, mean arrow speed varied by between 0.23% and 0.58% from one shot to the next, less than a tenth of the variation in the time between the clicker dropping and the release. In three archers from the Malaysian national squad, radar measured 197.9 ± 1.0, 219.4 ± 1.8 and 220.3 ± 0.9 kilometres an hour arrow by arrow at seventy metres, standard deviations of between a quarter and half a metre a second. In the laboratory an archer from the Japanese national team repeated her launch speed within 0.6%, against 0.3% for a compressed-air launcher.
That an expert archer repeats the arrow's speed from shot to shot almost as closely as a shooting machine is therefore reported: the same order of magnitude as the machine, but in the only direct comparison, on a single archer, her spread was twice as large. How much of an archer's vertical spread comes from this fraction of a metre a second has not been measured.
Speed and score
That leaves the second half of the question, whether a faster arrow brings points, and it is the one on which the measurements say least. In the German sample neither mean arrow speed nor its consistency entered the model that explained the score, which followed the regularity of the release instead. Three of the four associations between draw weight and score point the same way, but in elite archers the effect is small and the strongest correlation comes from a sample that mixes recurve and compound: what can be read from the claim that a heavier draw weight leads to higher scores, which is reported, is that draw weight and level go together, not that the pounds bring the points. More on this in the article on draw weight and score.
The other argument for speed is the wind. That a faster arrow drifts less in a crosswind is transferred, because only calculations say so, and on one condition: the speed has to come from a more powerful bow, not from a lighter arrow. In a model at seventy metres with a three metre per second crosswind, the same compound arrow drifts 175 millimetres when launched at 78.2 metres a second and 150 when launched at 86.5; with a recurve X10, on the other hand, the 140 grain point sends it off at 55.7 metres a second instead of 58.2 and drift falls from 261 to 245 millimetres, because the slower arrow is also the heavier one. The rest is in the article on which arrow drifts least in the wind.
What I take from this for the shooting line
These measurements do not give a speed to chase. They give an order of magnitude and a mechanism: if speed changes from shot to shot the arrow moves up or down, and according to the models even a fraction of a metre a second shifts the impact by a few centimetres at seventy metres. None of these findings comes from an experiment that changed an archer's arrow speed and measured the score afterwards, and I keep that in mind in the three things I take from them.
The first is that, if vertical spread worries me, I would look at whatever makes speed change from one arrow to the next before looking at speed itself. The machine tests point to draw length, since a release triggered by the clicker narrowed the vertical spread, and to arrow weight, with heavier arrows landing lower even for differences of a few tenths of a gram. These are tests without archers, but they show where to look: draw length, which I discuss in the article on the clicker, and a set of arrows as even in weight as possible.
The second concerns the chronograph. Since a commercial instrument can read a couple of percentage points below the reference and round to the whole metre per second, I would use it for comparisons on the same instrument, in the same light and in the same session, between two strings or two sets of arrows for example, rather than to set my number against another archer's.
The third concerns how to gain speed, if you want it. There are two routes, more draw weight or a lighter arrow, and they cost different things: a lighter arrow takes less energy from the bow and, in the wind calculations, drifts more; for extra draw weight what we saw above applies, namely that it goes with level without anyone having shown that it brings points on its own. How to choose a draw weight is covered in the article on choosing your draw weight; on the measurements we have, gaining a few metres a second is not in itself a reason to change your set-up.
Where the research stops
The question closest to the archer has not been tested: nobody has had the same archer shoot arrows launched at different speeds to see whether the score changes. The link between launch speed and vertical spread, too, has been measured with machines and with a single archer on film, never in a group of archers at seventy metres, and the difference between compound and recurve, clear in its direction, has not been quantified with the same draw weight and the same arrow.