That shake in the bow arm at full draw, when the aim should be holding still, is one of the first things an archer notices about themselves, and it is almost always treated as a flaw to hide. Some movement is unavoidable, because no muscle holds a load perfectly still; the measurements made on archers also show, though, that what we call tremor is not a single thing. It contains two movements: a slow sway that carries the sight around the target and runs mostly sideways, and a faster, smaller tremor that makes it shiver. In the only study to compare the two in this way, the slow part was the larger one in every archer measured.
The distinction changes the advice. While aiming, the movement tends to close down as the release approaches, and fatigue increases the bow's sideways sway; the fast tremor shows up in the signal of the muscles holding the draw, and at full draw, in the only study to measure it, more muscle activity did not go with less tremor. Its measurable counterpart is steadiness of force: at the same maximal strength, archers hold a steadier force than people who do not shoot, and among fifteen recurve archers those who held it most steadily scored higher. Below are the measurements, what they show and how far they go, and at the end what to do with them in training.
Two movements inside the same shake
The study that separated the two movements most clearly fixed an accelerometer to the bows of 16 archers aged 10 to 20, six shooting compound at 50 metres and ten shooting recurve at 30, and broke the movement during aiming down by frequency. The authors called the part between 1 and 10 Hz bow arm sway and the part between 10 and 20 Hz tremor. In all sixteen the slow part was larger than the fast part, and it was also the part that varied most from one archer to the next, with a spread of 47 points against 24 on the study's scale. It is a descriptive study, with no statistical tests and two bow types and two distances mixed together, and for now it is the only one to make this comparison.
Where sway ends and tremor begins is, in any case, the choice of whoever is measuring. In a Danish study of 15 recurve archers, presented at a conference in 2011, horizontal bow movement had a large share of its power below 1 Hz, meaning it was largely a slow drift, and it was not linked to the score; what was linked was vertical tremor between 3 and 8 Hz, which under the 2024 split would fall in the slow band. The figure sets the bands used by the different studies side by side: when you read how many points "tremor" costs, you need to know which movement is meant.
That the aim moves more horizontally than vertically is found by three independent studies: in 12 Taiwanese archers at 70 metres, with a laser pointer on the grip, it happened in ten out of twelve; in 7 international-level junior recurve archers, at 30 metres indoors, in all seven over the last second before the shot; in 4 elite Turkish women archers at 18 metres the point of aim travelled 13.0 centimetres horizontally across the target against 11.6 vertically. The ratio between the two directions does change from archer to archer, though, and in two of the twelve Taiwanese archers it was level or reversed. For anyone watching their own sight, this means a slow sideways movement is the norm and on its own does not point to a fault.
The body sways over the feet too, and there the studies do not even agree on the main direction. Over the whole draw and aim, two small studies of four elite archers each find more movement across the target; three more recent and larger studies, in the last second before the release, find more movement along the line to the target, for example 13.5 millimetres against 10.7 in 39 elite archers. The difference may lie in the time window measured, but nobody has compared the two windows in the same study.
The fast tremor: where it comes from
Tremor in the strict sense is what muscles produce when they hold a load: motor units switch on and off continuously to sustain the tension, and the result is a fine, fast oscillation. In archers it can be followed all the way to the bow. In seven university archers holding the draw to exhaustion, the tremor peak between 8 and 12 Hz in the electrical signal of the posterior deltoid moved in step with the oscillation of the bow limb, with a coherence between 0.81 and 0.96.
Where along the arm it starts has been measured by a single study, on the drawing arm. Ten beginners held a 26-pound recurve in six simulated positions, with motion sensors on the upper arm and the shoulder blade and electromyography of eight muscles: tremor was larger at the upper arm than at the shoulder blade, both between 8 and 12 Hz and between 3 and 7. The link with muscle activity was weak and changed sign: at partial draw more activity went with less tremor, at full draw partly with more, and the strongest correlation explained about a third of the variation. In the same experiment, with the drawing arm raised to 100 and 110 degrees, tremor was larger than at 90, while muscle activity stayed the same: a drawing elbow above shoulder height cost steadiness, not muscular effort. These were beginners, holding positions without shooting, so the finding points in a direction rather than setting a rule for a trained archer; for alignment itself, see the article on the drawing elbow.
For anyone shooting, the cautious reading is that the fast tremor does not go away with harder contraction: in those ten beginners, at full draw, more muscle activity did not go with less tremor, and in part went with more. It is a correlation, not a test in which anyone was asked to squeeze harder.
As for the tremor that reaches the bow, the only measurement of its spectrum comes from a Japanese engineering group that wanted to tune a damper to the person shooting. With accelerometers on the sight and three people seated with the bow at full draw, seated precisely to remove trunk and leg sway, the published spectrum shows two peaks, near 8 and 14 Hz, which stay in place over three repeated measurements of the same person while the amplitude changes each time. It is a single subject, not described as an archer, and in the example of the model fit the peaks sit at 4.6 and 12.8 Hz: the two frequencies describe that person's tremor rather than a value that holds for everyone, and they say nothing about what happens standing on the shooting line.
What reaches the target
The geometry explains why such small movements matter. The ten ring of an 80-centimetre face at 50 metres subtends 5.5 minutes of arc, and with a 70-centimetre arrow keeping the sight inside it means holding the bow steady to within about half a millimetre; from the other side, a sight one metre away that moves by 1.74 millimetres sends the arrow 122 millimetres off at 70 metres. These are calculations, not measurements. In archers we have the real scale of the movement: in 15 Turkish national team archers at 18 metres, the aiming point moved by about 8 centimetres on the target in the last three tenths of a second before release, and that range was linked to the score in only two of the four conditions tested.
That less movement means better arrows seems obvious, yet the studies divide: the results in favour are the more numerous, but there are null results as well, and each study measures something different and calls it bow stability. A clearer finding concerns timing. In 7 international-level juniors, at 30 metres indoors, the area in which the aim moves shrinks second by second: between 0.85 and 1.79 square points from three to two seconds before the shot, and between 0.30 and 0.59 in the last second, which was the smallest for every archer. In that last second the aim was also tighter on 10s than on arrows scoring 8 or less. For an archer this means that movement in the first seconds of aiming is normal, and that what matters more is how the aim settles before the release; this too comes from a single study, of seven archers at one distance.
Steady force, the other side of tremor
Tremor has a counterpart that can be measured: the ability to hold a force steady. In a Turkish comparison, 65 traditional archers and 65 sports science students held half of their maximal elbow flexion strength for thirty seconds while watching their own force curve on a screen. Maximal strength was the same in the two groups, 71.2 against 71.5 kilograms, but the archers held it more steadily: the coefficient of variation was 6.17% against 7.20%, and they stayed within 2% of the target for 80% of the time against 74%. The difference was clear in men and did not reach significance in women, who had on average four years of practice rather than six. The comparison was made once, on the biceps rather than at full draw, so it cannot say whether steadiness comes from shooting or leads people to it.
That steadiness matters for the score is suggested by the Danish study mentioned above. Its 15 recurve archers, from sub-elite to elite, had to hold a force equal to their own draw weight in the aiming position while watching its curve, and the more that force varied, the lower their mean score over 150 arrows at 18 metres: steadiness alone accounted for 45% of the differences, while maximal strength was not among the predictors of score. It is an association measured once in fifteen people and published as a conference abstract, and nobody has yet tried training steadiness to see whether scores rise.
The same study also touches on draw weight. Bow force averaged 196 newtons against a maximal pulling strength of 532, around 37%, and when the bow exceeded 40% of the maximum the score tended to fall, but without reaching statistical significance (p = 0.054): for the registry the 40% threshold has not been put to the test. The authors draw a suggestion from it, to take maximal strength into account when choosing draw weight, and for now it remains a suggestion.
Fatigue
Fatigue is the factor with the most direct measurements. In the seven university archers who held the draw to exhaustion, for 52 seconds on average, the frequency of the posterior deltoid's electrical signal fell in all of them, and the muscle tremor between 8 and 12 Hz showed up in the oscillation of the bow limb; the same signal also contained a slower oscillation, between 2 and 5 Hz, which the prolonged effort did not change. A year later, in 12 archers from the Italian federation, a protocol of six sets of ten 20-second holds at full draw increased the bow's sideways sway in all twelve, by 6% to 39% and 20% on average, while the centre of pressure under the feet did not shift. In those measurements, then, fatigue moves the bow and not the base. These are two small studies, with no control group.
The same twelve archers scored the same at 12 metres after the protocol as before, and the sway grew most in those with lower FITA scores: at that distance the score held up even with a bow that moved more. On form at the end of a session, it is mostly archers themselves who report. In a questionnaire of 59 university students who practise kyudo, 58% said that at the end of training they could still have shot, but not with good form: it is the practitioners' impression rather than a measurement, and it points the same way as the laboratory data. I have written about this at more length in the article on fatigue in archery.
How to reduce it
Before changing anything, notice when the tremor arrives. It is a rule of thumb from the field rather than a measurement, but it helps you get your bearings: if the arm shakes noticeably from the first few arrows, the first suspect is draw weight or basic technique; if the tremor builds in the final sets, fatigue; if it shows up mainly in competition, tension.
The first thing to check is still draw weight. Research has not fixed a threshold, but how steadily the shoulder holds your own draw weight goes with the score, and the Danish authors themselves point out that shoulder steadiness falls as the relative load rises. In my experience dropping a few pounds is often the single change that makes the biggest difference, and I cover it in the article on draw weight and your score.
The second is training. The only experiment that measured bow tremor while aiming after physical training was run on 20 university students who had never shot: side bridges with the trunk inclined at 40 degrees, five days a week for four weeks, and with a 20-pound bow tremor fell in the trained group and not in the control group. Whether the same work steadies the aim of a trained archer has not been measured with a bow in hand. Between maximal strength and endurance, the few studies on archers do not find holding capacity more important than strength: both count, and a specific strength programme makes sense if it trains both. The authors of the Danish study propose aiming exercises with visual feedback, precisely to train steadiness; it is a proposal, and its effect on scores has not been put to the test.
The third is equipment, which acts on what happens to the bow rather than on the arm. The same Japanese group built two dampers tuned to the tremor frequency of two people, at 9.30 and 8.70 Hz, and fitted to the bow they reduced the amplitude of the tremor by 35% and 80%, about half on average; the two people were not described as archers, the tremor was measured seated and nobody looked at the group on the target. In barebow, where weights on the riser take the place of stabilisers, seven club archers rating eleven set-ups blind preferred those that left the bow steadier: with two weights, the heavier in the centre, the bow moved 17 to 19 millimetres sideways after the shot, against 139 for the bare bow. This concerns behaviour after the shot, though, not tremor while aiming, and telling two set-ups apart by score would take, the authors calculate, between 900 and 1,050 arrows each. For the Olympic recurve the question runs through stabilisers.
That leaves the surrounding variables. On caffeine the studies on archers do not agree even about accuracy, and in seven elite clay target shooters, double-blind, it did not improve the score, while the shooters reported headache, anxiety and tremor. Sleep and recovery belong on the list because of fatigue: in my experience an archer who arrives at a session tired starts seeing the bow sway sooner, although nobody has measured this. And many archers say they shake more in competition than in training: it is common experience rather than a measured result, but it is why breathing before the shot belongs on this list.