What training load actually measures
Training load is a measure of the total physical stress a session, or a whole block of training, places on your body. It is built from two ingredients: volume, how much you did, and intensity, how hard you did it. A long, easy run and a short, brutal interval session can land at a similar training load number even though they feel nothing alike, because one trades duration for effort and the other trades effort for duration.
The most common way to quantify a single session is session RPE load: multiply the session's duration in minutes by your rate of perceived exertion on a 0-10 scale. A 45-minute session at an RPE of 7 produces a load of 315. Strength training more often uses volume load instead, sets multiplied by reps multiplied by weight, which captures mechanical work directly rather than a subjective effort rating, and it is the same number an AI strength training plan uses to decide when to add weight versus a set. Endurance training sometimes swaps in TRIMP or a pace-and-heart-rate-based formula, which weights time spent at higher heart rate zones more heavily than easy zones, and it is the same underlying math behind Strava's fitness and freshness chart, which is really acute and chronic load under a different name.
None of these single-session numbers mean much on their own. A load of 315 tells you almost nothing in isolation, it only becomes useful once you can compare it against what you have been doing recently and what you have adapted to over the last month. That is the entire reason training load is tracked as a rolling trend rather than a one-off score.
Acute load, chronic load, and the ACWR
Acute load is a short-term rolling average, most commonly 7 days, that reflects what you have been doing recently. Chronic load is a longer rolling average, most commonly 28 days, that reflects the baseline of training your body has actually adapted to. Chronic load moves slowly by design, a single hard week barely nudges it, which is exactly what makes it useful as a stable reference point for the more volatile acute number.
The acute:chronic workload ratio, or ACWR, is simply acute load divided by chronic load. Say your chronic load, averaged daily over the last 28 days, works out to 300, and your acute load, averaged daily over the last 7 days, jumps to 450 because you added two extra hard sessions. That gives an ACWR of 450 / 300, or 1.5, meaning this week's training is running 50 percent above what your body has recently adapted to.
Sports science research, drawing largely on team-sport and endurance injury-surveillance studies, generally places the lower-risk ACWR band between about 0.8 and 1.3. Injury risk climbs as the ratio pushes past 1.5, and the risk is highest when a big spike follows a period of low chronic load, since there has been little time to build tissue tolerance. This is a risk signal built from population-level patterns, not a guarantee of injury or safety at any specific number, and it works best as one input alongside how you actually feel.
Honest ranges by age, sex, and goal
The 0.8 to 1.3 ACWR band holds up broadly across age and sex, because the ratio is self-relative. It compares your training this week to your own baseline, not to a fixed number derived from other people, so a 60-year-old and a 25-year-old can both sit safely at an ACWR of 1.0 despite having very different chronic loads in absolute terms.
What genuinely differs by age, training history, and goal is how fast it is safe to raise chronic load itself, not the ratio band. Beginners, older adults, and anyone returning from injury generally do best building chronic load in small, gradual increments, roughly single-digit percentage increases week over week, with more recovery days placed between any load spikes. Trained lifters building toward a hypertrophy block, and runners building toward a race, can typically tolerate a faster ramp and a higher absolute chronic load, provided sleep and other recovery markers hold steady through the increase.
Absolute load numbers are not directly comparable between people or even between your own sessions if the calculation method changes, since volume load, session RPE load, and TRIMP all use different units. The relative framework, how this week compares to your own last month, is what stays honest and comparable over time, regardless of which method produced the underlying numbers.
How to read your own training load trend
The practical signal to watch for is acute load spiking well above chronic, pushing the ACWR past roughly 1.3 to 1.5, especially after a missed week, the start of a new program, or the build-up phase of a race plan. That pattern, a sudden jump in recent training against a baseline that has not caught up yet, is the exact scenario the research flags most consistently.
The ratio is more useful paired with subjective and physiological cues than read in isolation: persistent soreness, poor sleep quality, or a resting heart rate that is running higher than usual alongside a rising ACWR is a stronger combined signal than any of those on their own. If you wear a device that already scores daily strain or recovery, such as Whoop, reading that number next to your logged training load catches spikes the workout history alone would miss. A flat or slowly declining chronic load over several weeks, on the other hand, usually means detraining, easing off too much, rather than a sign of anything protective.
Reading the trend well means checking it periodically rather than reacting to a single day's number, and it works best when it sits next to the rest of your training picture, per-muscle fatigue, session intensity, sleep, and recovery, instead of standing alone as one abstract ratio you have to interpret from memory.
See your own acute and chronic load, tracked automatically
Wellness Project turns every logged workout, run, and recovery day into a rolling acute and chronic training load trend, with your specialist coaches reading it alongside sleep, HRV, and heart rate. No spreadsheet, no manual RPE log. Free during early access on iOS, Android, and web.