No study has ever used a jump test to catch a running injury before it starts. That’s the honest answer on reactive strength index and injury prediction. The test still tells you something real: how tired your tendons are, right now, today.
Reactive strength index (RSI) comes from jumping off a box, landing, and springing straight back up. Coaches built it to measure a sprinter’s explosive power. A wave of running content now claims it can flag your next overuse injury before symptoms show. That claim runs well ahead of the science.
What Reactive Strength Index Actually Measures
RSI measures how well your legs handle a fast stretch, then a release. You drop off a box, land, and jump back up immediately.
RSI = Jump Height (m) / Ground Contact Time (s)
In short: a higher number means you sprang back up fast and high. A lower number means you sank into the landing and pushed off slow.
There’s a friendlier version for runners without a coach or a force plate. RSImod swaps ground contact time for time-to-takeoff on a plain countermovement jump, the crouch-and-jump you’d do standing on the floor.
RSImod = Jump Height (m) / Time-to-Takeoff (s)
No box. No landing technique to get wrong. For a runner self-testing at home, RSImod is usually the better starting point.
Same idea. Easier setup.
What the Evidence Actually Links Reactive Strength To
Here’s where the science holds up well. Runners and other athletes with Achilles tendinopathy show real, measured deficits in reactive strength. A 2019 systematic review and meta-analysis in the Journal of Athletic Training pooled the tendinopathy research (McAuliffe et al., 2019). Patients with a painful Achilles jumped 16% to 35% lower on reactive, hopping-based strength tests than their healthy side.
Translation: a sore Achilles doesn’t just hurt. It measurably weakens your spring.
Notice the tense, though. Every one of those patients already had a painful tendon when researchers tested them. That’s a snapshot of people who are already hurt, not a forecast of who’s about to get hurt.
A sick tendon shows up on a jump test. A healthy one that tests low isn’t the same thing.
Normative RSI Values: Sprinters, Team Athletes, and What We Actually Know About Runners
Every popular RSI chart online is built for sprinters or team-sport athletes. Almost none cover distance runners.
| Population | Test type | Typical RSI | Source |
|---|---|---|---|
| Elite 100m sprinters | Drop jump | 3.0-4.0+ | Practitioner data (Flanagan, via ScienceForSport) |
| General athletic “good” level | Drop jump | Above 2.0 | Practitioner data |
| Recreational athletes | Drop jump | 1.0-1.5 | Practitioner data |
| Amateur distance runners, under 30 | Drop jump (20cm) | 2.09 (±0.50) | Jaén-Carrillo et al., 2021 |
| Amateur distance runners, 30-50 | Drop jump (20cm) | 1.81 (±0.43) | Jaén-Carrillo et al., 2021 |
| NCAA D-I mixed-sport males | CMJ, RSImod | 0.419 m/s (median) | Sole et al., 2018 |
| NCAA D-I mixed-sport females | CMJ, RSImod | 0.308 m/s (median) | Sole et al., 2018 |
Only one study has published RSI numbers for distance runners tested as their own group. It tested 59 amateur male runners on a 20cm drop jump. Runners under 30 averaged 2.09. Runners 30 to 50 averaged 1.81, a drop that didn’t reach statistical significance (p=0.154) but showed a moderate real-world effect (0.63).
In short: your RSI likely drifts down with age, like most power measures. Nobody has pinned down exactly how fast, or what that means for injury.
One more catch. Jump technique alone changes the number by 20% to 35% at a matched drop height, in the same athlete (Struzik et al., 2016). Box height moves it too. Test off a 30cm box one week and a 45cm box the next, and you’ll see a “change” that’s really just a different test.
Keep your setup identical or the trend means nothing.
The Gap Nobody Mentions: RSI Doesn’t Predict Injury in Runners
Here’s the claim you’ll see on training blogs: test your RSI, catch a weak spot, stop an injury before it happens. No prospective study backs that up. Not one trial has tracked healthy runners’ RSI over months and then checked who got hurt.
The tendinopathy numbers above come from people who already had a sore tendon when tested. That’s the gap between “sick tendons test weak” and “weak-testing tendons get sick.” Only the first one has real evidence behind it.
Think of RSI like a smoke detector installed after the fire. It confirms the room got hot. It doesn’t warn you before the match strikes.
There’s a deeper problem specific to running. A study built to check whether reactive strength connects to running mechanics found it barely does. Jaén-Carrillo and colleagues (2021) compared RSI against leg stiffness during submaximal running, the actual spring-like quality your leg shows mid-stride. The correlation came in under 0.12. Statistically, that’s no relationship at all.
Translation: how well you bounce off a box tells you almost nothing about how your leg behaves at mile 10 of an easy run.
A jump test and a running stride aren’t the same movement.
The closest anyone has come used ultrasound, not a jump test, in teenage volleyball and basketball players (Mersmann et al., 2023). High tendon strain meant 2.3 times the odds of tendon pain later. That’s a real before-the-injury finding. It just isn’t built from RSI, and it isn’t built from runners.
How Much RSI Change Is Real? Reliability, Noise, and the Smallest Worthwhile Change
RSI itself is a reliable test. A 2026 study in Sports measured drop-jump RSI test-retest reliability at a coefficient of variation (CV) of 6.2% and an ICC of 0.89 (Clancy et al., 2026). Test again next week under matched conditions, and you’d expect a similar score.
In short: the test doesn’t lie to you randomly. It’s consistent.
Consistent isn’t the same as sensitive. The same study compared typical error, how much a score naturally bounces test to test, against the smallest change actually worth caring about.
| Test | CV% | ICC | Typical error | Smallest worthwhile change | Verdict |
|---|---|---|---|---|---|
| Drop jump RSI | 6.2% | 0.89 | 0.09 | 0.05 | Marginal for small changes |
| 10-5 repeated jump test | 5.2% | 0.89 | 0.08 | 0.04 | Marginal for small, good for moderate changes |
Typical error came in larger than the smallest worthwhile change on both tests. Researchers rated a single test only “marginal” for catching a small, real change.
A one-point drop in your RSI might be real. It might be Tuesday.
Bigger swings tell a clearer story. Moderate or larger changes come through the noise fine.
A Single Low Reading After a Long Run Doesn’t Mean What You Think
Fatigue alone can crash your number for a day. Elite trail runners tested before and after the 2023 World Trail Running Championship, a 45.2km mountain race, showed a real drop in RSI right after finishing (Jaén-Carrillo et al., 2025). Jump height barely moved. RSI did, a large effect in a small group of 8 athletes (d=0.862, p=0.045).
Translation: your legs can look the same in the mirror and still spring less after a hard effort. RSI caught something jump height alone missed.
Take a runner I’ll call Priya, 34, training for her first 50-mile trail race. She tested RSImod with a phone app every Monday morning, rested, never after a hard session. For a month, her number sat flat around 0.42. In week 6, the day after her longest run yet, 22 miles with real climbing, it read 0.31. A week later, on an easier week, it bounced back to 0.40.
Nothing was wrong with her tendons. She was reading a tired Tuesday as a broken Wednesday.
One low number after a hard week is noise, not a diagnosis.
How to Self-Test RSI at Home With a Phone
You don’t need a lab. Apps like My Jump 2 measure jump height and contact time from slow-motion video. Validity studies show both correlate above 0.98 with force-plate readings.
A few rules keep your number trustworthy:
- Pick RSImod (a plain countermovement jump) over a box drop jump if nobody’s coached your landing technique.
- Test rested, at the same time of day. First thing in the morning, before a run, works well.
- Keep drop height and jump style identical every single time.
- Do 3-5 jumps and keep the best one.
- Retest every 1-2 weeks, not daily. Daily fatigue swings will bury a real trend before it shows up.
If you want to build the tendon spring RSI measures, not just test it, five minutes of daily hopping has real trial evidence behind it.
Same setup, same time, same effort. That’s what makes a trend worth trusting.
Tracking RSI in Context, Not as an Isolated Score
AthleteOS lets you log a periodic RSI or RSImod result alongside your training load and workout history. A dropping number stops being an isolated score with no explanation. You’d see something like reactive strength down 18% during the three weeks your weekly mileage jumped 40%.
Because AthleteOS tracks training load week by week, it can flag when a jump-test drop lines up with a mileage spike or a block of hill repeats, instead of leaving you to guess between normal fatigue and a real trend. See how AthleteOS tracks your training data over time.
RSI shares its biggest lesson with the drift ratio your long runs already generate: one bad number is noise, a repeated trend is signal. If you’re building mileage, the first 8 weeks carry outsized injury risk, and that’s the load context an RSI drop needs. If a jump test flags one leg over the other, the leg-strength asymmetry research covers which gaps are worth chasing. Pairing jump testing with heavier lifting? The concurrent-training research covers scheduling both.
Test it. Trend it. Don’t let one number tell you a story the evidence hasn’t earned yet.