The horse’s weight, the rider’s and the tack’s, and the load as a share of the horse — then what each study found at that share, with how many horses it used. The 20% everybody quotes is one of them.
Rider and tack, as a share of the horse
17.3% of his weight
What each study found at that share
| Share | Total load | Rider |
|---|
A percentage is not a verdict on a horse or a rider. The studies here used a handful of horses each and do not agree, and fitness, conformation, saddle fit, the rider’s balance and the work being asked all matter as much as the scale. If you are unsure whether a horse suits a rider, ask a vet or an experienced instructor who can see them both.
Studies checked 2026-09-25.
Quick answer: the number everybody quotes is 20% of the horse’s body weight, rider and tack together. For a 1,100 lb horse that is 220 lb, so about 190 lb of rider once a 30 lb saddle is on. That figure comes from one study of eight light riding horses (Powell, 2008). The studies since point both ways. A 2020 UK pilot stopped every test with a rider at about 15–18% of the horse’s weight. A 2020 Danish study saw no measurable change at 15–23% in a light dressage test. The honest answer is a share, checked against each of them.
How to use it
Put in the horse’s weight, the rider’s and the tack’s. The tack means the saddle, pad and bridle, and anything in the saddle bags. The answer is the load as a percentage of the horse. Under it, the page shows what each study found at that percentage, how many horses it used, and the biggest rider that 15%, 20% and 25% would allow with the tack you entered.
If you do not know what the horse weighs, measure him with a tape and the free horse weight calculator. A guess that is 100 lb out moves the answer by nearly two percentage points.
Where the 20% rule comes from
In 2008, Powell’s team at Ohio State’s Agricultural Technical Institute rode eight light riding horses through an exercise test carrying 15%, 20%, 25% and 30% of their body weight. Soreness tended to rise at 25% and rose significantly at 30%. The horses with wider loins were less sore. The team concluded that a total load of rider, saddle and equipment “not in excess of 20%” was quite acceptable for that kind of horse. It is a careful study, and it is eight horses of one type doing one test.
What came after
In 2020, Sue Dyson’s team at the UK’s Animal Health Trust rode six horses in the same 30-minute dressage test with four riders of different weights. Every test with the heavy rider, on average 16.7% of the horse’s weight, was stopped because of temporary lameness and behaviour consistent with pain, as were the very heavy rider’s. Dyson herself warned that this “does not mean that this is a reliable cut-off ratio”. It was a pilot study.
The same year, a Danish team at Aarhus University added 15% and 25% to the usual riders of 20 horses. At the heaviest, the riders were 15–23% of their horse’s weight. The team found no measurable change in stress hormones, heart rate or gait in a light dressage test, and said the result “should not be extrapolated” to harder work. In 2025, a Swedish and Icelandic team went looking for a test that could predict one horse’s carrying capacity, and did not find one.
Put together, the studies say something more useful than any single rule. The same share of body weight can be fine in a light test and too much in a harder one. The horse’s build, fitness and work, the fit of the saddle and the rider’s balance all move the line.