Kettlebell swing

Movement

The swing is the exercise studied most often in kettlebell research¹. The bell, its styles and a 2019 review of the field are on the kettlebell training page¹.

Hinge or squat

A 2019 scoping review, which maps what research exists on a topic, describes three swing patterns¹:

the hip hinge or “Russian” swing, to chest height, in a pattern like a deadlift¹;
the two-handed overhead or “American” swing, linked with a squatting motion and with CrossFit¹;
a double-knee-bend pattern, linked with kettlebell sport¹.

The same review reports that experts in the two-handed hardstyle swing hinge at the hips rather than squat¹. At the bottom, they had about 20 degrees more hip bend and 15 degrees less knee bend than novices¹.

Shoulder height or overhead

In 15 healthy adults, the overhead swing took longer per cycle than the shoulder-height swing¹¹. The same study found peak hip joint moments, the turning effect of a force around a joint, almost identical in the two styles¹¹.

In 18 women experienced with the overhead swing, the hip made the largest contribution and the knee the second largest¹². Those shares did not change across 8-, 12- and 16-kg bells¹². In 15 women with at least six months of swing experience, the overhead swing produced more hip output than the shoulder-height swing³:

How much higher in the overhead swing (%). Bars start at zero; the table below gives each value with its source.

Measure at the hip	How much higher in the overhead swing
Peak power	About 42%³
Work	About 22%³
Peak velocity	About 10%³

What it loads

The buttock and hamstring muscles

In 23 healthy college-age adults, two-handed and one-handed swings averaged about 75% of a maximal voluntary isometric contraction in the gluteus maximus¹³. That is a muscle’s hardest effort while the joint stays still. The same study measured about 55% in the gluteus medius and 79% in the biceps femoris, the outer hamstring¹³. In the same study, biceps femoris was the first of the measured hip muscles to switch on¹³. A longer resting length of the hip flexors, the muscles that bend the hip, was moderately linked with more gluteus maximus activity¹³.

Several studies compared the inner and outer hamstrings^2,14,15:

Who was studied	What was found
14 people, three swing styles	Inner hamstring activity was higher than biceps femoris activity in each of the three styles²
16 elite female handball and soccer players with no history of knee injury	The swing and the Romanian deadlift drew 17 to 22% more from the semitendinosus, an inner hamstring, than from biceps femoris¹⁴
A 2025 meta-analysis (a study pooling earlier studies)	Semitendinosus activity was greater than biceps femoris activity in swings¹⁵

The same meta-analysis cautioned that differences in study methods make it hard to say firmly whether any exercise favours either muscle¹⁵.

The trunk

A 2019 study of the American swing tested 15 resistance-trained men with a 14-kg bell¹⁶. The two-armed swing drew 40 to 59% more rectus abdominis activity than the one-armed swing¹⁶. External oblique activity did not differ¹⁶. In the one-armed swing, erector spinae activity was 14 to 25% higher on the side away from the working arm¹⁶. The study had no funding from companies, manufacturers or outside bodies¹⁶.

A systematic review, a set, checkable search of the research on a question, found three free-weight studies of the internal oblique¹⁷. Its highest value came from a swing with a forceful finish¹⁷. Two studies looked at fatigue in the low-back muscles^18,19:

In 8 active men, one set of swings lowered the peak torque, or turning force, of the low-back extensors¹⁸. It lowered it less than an exercise isolating those muscles did¹⁸.
A randomised trial studied 41 adults aged 18 to 45 with no recent low back pain¹⁹. High-intensity interval swings caused no significant difference in erector spinae fatigue, against a warm-up alone¹⁹. Its registry record names the University of Central Florida as lead sponsor²⁰.

The shoulders and back

In 20 healthy men, shoulder muscle activity was largely comparable in two-hand kettlebell and dumbbell swings²¹. The one-hand kettlebell swing drew more from some muscles, with the lower trapezius highest²¹. The differences between exercises were modest²¹. A study of footballers also measured latissimus dorsi activity during the swing²².

Forces on the body and spine

In male amateur lifters, peak ground reaction force, the push of the ground back up on the body, was about twice body weight¹. In 7 men, the swing produced a backward shear force at the low back, opposite in direction to a traditional lift²³. A shear force pushes two surfaces to slide across each other.

In 30 active college-age adults doing hip hinge swings, bells of 10 to 20% of body mass were compared²⁴. Heavier bells raised joint moments at the low back, hip and ankle²⁴. In 18 women doing the overhead swing, going from 8 to 16 kg raised hip loading most, then ankle, with little change at the knee¹².

In 16 experienced men, 10 rounds of 30 seconds of swings were each followed by 30 seconds of rest²⁵. Within each round, swings slowed and ground reaction force rose, while hip power fell²⁵. The authors call the larger force at the end of rounds a potential concern for injury²⁵.

The heart

In 14 young adults, the same number of swings was done in two ways²⁶. Intervals with 10-second rests gave higher oxygen use and peak heart rate than 4 sets with 90-second rests²⁶.

How the body adapts

The 2019 review on the kettlebell training page rated confidence in reported kettlebell effects as low to very low¹. It judged the data too thin to strongly support claims of better performance or fitness¹. It read this as a lack of data, not as evidence of no effect¹.

A 2024 narrative review looked at kettlebell training in athletes, not the swing alone²⁷. It searched four databases from 2013 to 2023 for trials with a control or comparison group²⁷. Its authors conclude that their synthesis suggests kettlebell training improved strength and power in athletes²⁷.

Training studies of the swing alone

These studies trained the swing on its own^4,5,6,7,8,9. Three of them placed people in groups by ability or convenience, not by chance^4,5,6. In the table, 1RM is the heaviest load a person can lift once:

Who trained, and for how long	What was found
28 men from high-intensity functional training, 6 weeks	Deadlift 1RM rose 7.8 to 8.7% and jump height 8.7 to 10.1% with a fixed or a stepped bell. Only the stepped group gained deadlift endurance, by 23.8%⁷
18 recreationally active women, 8 weeks, against stiff-leg deadlifts	Neither group changed its 30-m sprint; vertical jump improved across both groups⁴
30 trained US Air Force personnel (20 finished), 10 weeks, a pilot study	1.5-mile run time did not change; 40-yard dash improved within the swing groups, but not between groups⁵
18 young female gymnasts, about 13 years old, 4 weeks	Fitness on a cycling test rose in the 9 who added swings, not in the 9 on regular training⁶
21 healthy men, 6 weeks, randomly assigned against jump squats	Half-squat strength rose 9.8% and vertical jump 19.8%, with no significant difference between the two kinds of training⁸
28 men and 25 women, 12 weeks	Deadlift 1RM and jump height rose in both sexes; men gained significantly more power⁹

The gymnasts were placed in groups by age, space and convenience of time⁶.

Studies of the back and of past injury

A randomised trial of 60 healthy adults measured pressure pain thresholds, the pressure at which pressing on a spot starts to hurt²⁸. Swings after a warm-up changed them over the low-back and piriformis (deep buttock) muscles²⁸. The change favoured the swing group over warm-up alone, measured by an assessor who did not know the groups²⁸. The trial’s registry record lists the University of Central Florida as lead sponsor²⁹.

In 42 male football players, those with a hamstring strain in the past 6 months showed lower activity during the swing²². Biceps femoris activity was 28.7% lower and gluteus maximus 24.3% lower than in uninjured players²². The study looked at each player once, so it cannot show whether the lower activity came before or after the injury²².

Progressive overload

A 2026 position stand describes progressive overload as the need to raise the stimulus on the muscle throughout a programme³⁰. Kettlebell trials set or raised the load in different ways^4,6,7,31:

Who trained	How the programme was set or progressed
Men from functional training	12 sets of 30 s on, 30 s off, twice weekly. One group kept 20 kg; one used 16, 20, then 24 kg, two weeks each⁷
Recreationally active women	Started near 9.1 kg, with a possible 2.3-kg rise after week 3 and again at week 6 if form held⁴
Young female gymnasts	12 rounds of 30 s on and 30 s off, 3 times a week, with a bell a quarter of body weight⁶
Adults aged 59 to 79 (the BELL trial)	3 months of progressive hardstyle training³¹

Bell weight and swing height

Kettlebells now range from 2 to 92 kg, but above 56 kg they come in 12-kg steps³². The authors of a 2023 study conclude that changing how high the bell is swung can make the same bell harder or easier³². In 15 men swinging 56 kg, swing height changed impulse, which is force added up over time, and the work done³². It did not change how long either phase lasted³².

In 15 experienced women, going from a 12- to a 16-kg bell brought fewer and smaller changes than switching to the overhead swing³. The authors suggest a heavier bell may be considered before the overhead style³. In 10 strength-trained men with total work kept equal, a 16-kg bell gave higher heart rate than an 8-kg bell³³. It also gave higher levels of two hormones, cortisol and testosterone³³.

Injury reports

One case report describes a 33-year-old sailor who developed back pain about 24 hours after a high-intensity swing workout¹⁰. He had rhabdomyolysis, a rapid breakdown of muscle tissue¹⁰. He also had compartment syndrome, a rise in pressure inside a closed space of muscle, in the muscles beside the low spine¹⁰. It needed an operation (fasciotomy)¹⁰.

The other figures cover kettlebells in general, not the swing^1,34. In 15,348 emergency department cases linked to resistance training, kettlebells were involved in 0.80% of injuries in women and 0.10% in men³⁴. A 2019 review found no adverse events reported in kettlebell trials, but could not say whether that reflected the training or limited reporting¹. The same review found no data suggesting that a swing done against the prescribed hardstyle pattern carries more risk of harm¹. Among novices, reports of delayed soreness and bruising where the bell hits the forearm are not unusual¹.

Connections

The muscles it works

In the studies above, the swing worked the gluteus maximus, gluteus medius and hamstrings¹³. It worked the erector spinae, the long muscles beside the spine, and the belly’s rectus abdominis, external oblique and internal oblique^16,17,23. It placed its own demands on the vastus lateralis, an outer thigh muscle of the quadriceps³⁵. In the back, studies measured the trapezius and latissimus dorsi^21,22.

The swing loads the low back of the spine with a backward shear in one study of 7 men²³. The barbell deadlift produces a forward shear there instead¹. In 16 men, swings produced more power than back squats, though force tended to be higher in squats³⁶. Two trials compared kettlebell training with jump training, covered under plyometric training^8,37.