Barbell velocity describes how quickly the bar moves, usually in meters per second (m/s). In strength training, a device will often report one number for the lifting, or concentric, part of a rep.
That number needs a label. A reading of 0.50 m/s means something different if it’s the average speed across the lift or the fastest instant in it.
Mean, peak, and propulsive velocity
Mean concentric velocity is the average velocity across the lifting phase. It includes the slower portions as well as the faster ones. When someone talks about a rep’s average bar speed, this is usually what they mean.
Peak velocity is the highest velocity reached at one instant. A rep can have a quick burst and a long sticking point, so peak speed won’t tell the same story as the average.
Mean propulsive velocity averages the part of the lift classified as propulsive, leaving out the braking phase. It appears often in VBT research. It’s a different calculation from averaging the entire concentric phase.
Researchers have compared these as separate variables, including in the bench press throw. Which reading is useful depends on the movement and the question being asked; the labels aren’t interchangeable. See García-Ramos and colleagues’ comparison.
Check which metric your device reports before comparing it with a paper, a chart, or another app.
What counts as a good bar speed?
There isn’t one speed that makes a rep good. A heavy deadlift and a light bench press aren’t supposed to move alike.
The useful comparison is your own performance on the same exercise, at the same load, with similar technique and effort. Research on powerlifters found relationships between velocity, relative load, and perceived effort, and recommended individual load-velocity profiles for prescription. Read Helms and colleagues’ study.
A load-velocity profile is a record of how fast you move different weights on a particular lift. It gives you a personal reference instead of asking you to fit somebody else’s chart. A profile for your squat doesn’t double as one for your bench.
Make the comparison fair
Say a hypothetical set has an average lifting speed of 0.45 m/s one week and 0.50 m/s the next. That’s faster, but the number alone doesn’t explain why.
Before calling it progress, check a few ordinary things:
- Was the load the same, including the bar?
- Did you use the same range of motion and pauses?
- Were you trying to lift with the same intent?
- Was the device in the same position, reporting the same metric?
Those checks matter more than the number of decimal places on the screen. Repeated changes under similar conditions are more useful than one unusually fast or slow rep.
Bar speed and rep speed aren’t always the same thing
A device attached to a bar measures or estimates the bar’s movement. A sensor on your wrist follows your wrist. On some exercises those move together; on others, grip, wrist rotation, and the movement path make the relationship less direct.
Treat the attachment position as part of the measurement. Don’t compare a wrist reading with a bar-mounted reading as though only the location of the sensor changed. A study comparing wearables and a phone app against a linear transducer illustrates why validation needs to account for the device and setup. Read the measurement study.
How Strive reports it
The Strive app shows the mean speed of each rep in m/s, labelled as such, along with the best rep so far in the set. Every set keeps its rep-by-rep speeds, so you can look back at exactly how it moved instead of a single summary number.
Because the puck logs your sets automatically, the fair comparisons above get easier: the same exercise, its weight, and its rep speeds are already recorded together, session after session. That history is what turns a single reading into something you can actually compare.
What to look at next
Once you know what your speed metric means, you can compare reps within a set. The drop from the fastest rep is velocity loss, and the next guide includes a calculator so you can work through an example.
For the bigger picture of how people use these readings, start with velocity-based training.