How to Calculate Your ACWR by Hand (Including Your Lifting)

August 27, 2026 · 6 min read

The acute:chronic workload ratio has a reputation for being complicated. It is one division. What is actually hard is getting two honest numbers to divide, especially if your week contains both running and barbells.

Here is the whole thing, then the parts that take judgement.

The formula

ACWR = this week's load / the average of the last four weeks' load.

That is it. The numerator is your acute workload, usually seven days. The denominator is your chronic workload, usually 28 days expressed as a weekly average, so the two numbers are on the same scale and the ratio comes out near 1.

A worked example. Say your last four weeks of load scored 420, 460, 400 and 480.

Chronic: 420 + 460 + 400 + 480 = 1760, divided by 4 = 440
This week you score 550
ACWR: 550 / 440 = 1.25

You did a quarter more this week than your recent normal.

Getting a number for the week

The formula is trivial; the input is where people get stuck. You need one number per week that represents everything you did, and there are three common ways to produce it.

Session RPE is the most portable and needs no hardware at all. Multiply the session's duration in minutes by how hard it felt on a 1 to 10 scale, rated about half an hour after you finish rather than in the moment. A 60 minute session at a 7 scores 420. Add every session in the week. Foster's group proposed this method precisely because it works across sports that share no common unit, and it is the reason it can price a squat session and a long run on the same scale. There is more on how it behaves across a mixed week in combining Strava and Hevy into one training load.

Heart-rate TRIMP is better for the cardio side if you wear a monitor, because it responds to how hard the session actually was rather than how hard you remember it being. It weights time spent at higher heart rates disproportionately, which is why a 40 minute threshold run can outscore a 90 minute easy one. It says nothing at all about a heavy set of five, since your heart rate barely moves. The comparison between the methods is worked through in TRIMP vs power vs RPE.

Tonnage handles the strength side: sets times reps times weight. Straightforward to compute and completely incomparable to a TRIMP score without a conversion factor you have chosen yourself. That conversion is a modelling decision, not a measurement, and anyone who tells you there is a correct constant is overstating what is known.

The practical advice: pick one method and use it for everything. A ratio built from sRPE across all sessions is more useful than a ratio built from TRIMP for runs and tonnage for lifts glued together with a fudge factor. The ratio compares this week to your own recent weeks, so consistency of method matters far more than which method you picked. Mixing methods mid-block is the single most common way to produce a number that means nothing.

Reading the result

Broadly, a ratio near 1.0 means this week resembles your recent normal. Below about 0.8 you are doing meaningfully less than you have been. Above about 1.3 you are doing meaningfully more.

Those are the bands used across the research literature, and they are also the point where honesty is required. The 0.8 to 1.3 range comes from Gabbett's work on training loads in team sports, and it has been argued about ever since. Later analyses have criticised the ratio on statistical grounds, including how the rolling averages are constructed and how much the choice of 7 and 28 days drives the result. The bands are a convention that proved useful, not a constant that was measured. Treating 1.31 as different in kind from 1.29 is reading precision into a number that does not have it. The fuller version of that argument, including what the critics get right, is in the honest case for and against ACWR.

What the ratio is good for is noticing the shape of a change you would otherwise miss: a week that felt fine because each individual session felt fine, but which added up to a third more than your base. It is a description of your ramp rate, not a verdict on it.

The part that catches hybrid athletes

Run the calculation twice, once on your running load alone and once on everything.

For a lot of people training two sides, the two ratios disagree. Running volume held steady, so the running number sits near 1.0 and looks unremarkable. Meanwhile a new lifting block added three sessions a week, and the combined number is 1.4. Both numbers are correct. Only one of them describes the week your body actually had.

This is the specific failure mode of tracking each side in its own app. Your watch computes a ratio from the half of the week it can see. Your gym app computes nothing of the kind. Neither of them is wrong, and neither of them mentions the other half. There is more on why the watch behaves this way in why your watch ignores your lifting.

The science FAQ sets out the papers behind each piece if you want to read the sources rather than take our word for the summary.

Doing it in a spreadsheet

If you want to run this yourself, the whole thing is five columns:

Week — the week ending date
Load — your one number for that week
Chronic — a four-week rolling average of the Load column
ACWR — Load divided by Chronic
Band — Ramping down below 0.8, Steady from 0.8 to 1.3, Sharp ramp from 1.3 to 1.5, Spike above 1.5

Two caveats on the spreadsheet version. You need at least four weeks of history before the chronic column means anything, so the first month of numbers will be noisy and should be ignored. And a week where you were ill or travelling drags the chronic average down for the following month, which makes an ordinary return to training read as a sharp ramp. That is arguably the ratio working correctly, but it is worth knowing before you react to it.

If you want the full version of this, including how to score the lifting side and put both on one scale, there is a walkthrough of building the whole training load spreadsheet with the formulas in it.

Or skip the spreadsheet

LOAD does this arithmetic on a week that already includes both sides. It reads heart rate from Strava or Polar and tonnage from Hevy or a manual log, scores them into one weekly number, and tracks the ratio across the combined figure. Connecting Strava backfills twelve weeks, so the chronic average is populated from the first session rather than a month later.

It is free for individual athletes, with no card and no paid tier. But the spreadsheet above genuinely works, and if you would rather own the arithmetic yourself, own it.