Your one-rep max is the heaviest weight you can move once with good form. Almost nobody needs to test it, and almost every strength plan is written as a percentage of it, which is why estimating it from a set you already did is the useful shortcut.
A set of 10 with four reps left in the tank is not the same as a set of 10 you could not extend. Prediction equations assume you stopped because you had to. If you feed them a comfortable set, the max comes out too low, and every working weight built on it comes out too light.
Rate of perceived exertion, on the scale with reps in reserve, turns an honest set into the number the equation needs. RPE 8 means two reps left, so a set of 5 at RPE 8 behaves like a set of 7 to failure. The calculator does that correction for you.
Epley and Brzycki disagree slightly, and the gap between them is the real precision of any estimate. Showing one number to the decimal would be pretending. The panel shows the range instead.
Every number on this page comes from a published equation that is named in the panel and cited at the bottom. If a calculator will not tell you which equation it used, it is not telling you what its number means.
A single target weight is the thing wrong with most strength calculators for cyclists. The load that builds force in November is the wrong load in June, because by June the job of the gym is to not cost you anything on the bike.
The four phases below are the plan we prescribe. The calculator turns each one into your numbers, from the same prescription the full guide publishes, so the two pages cannot drift apart.
| Phase | What it builds | Typical session |
|---|---|---|
| Anatomical adaptation | Movement quality and tissue tolerance | 48 min, RPE 6 |
| Maximum strength | Force production | 49 min, RPE 8, long rests |
| General strength | Holding the gains alongside riding | 50 min, RPE 6 to 7 |
| Maintenance | Keeping it when cycling load rises | 45 min, reduced volume |
Sets, reps, rest and RPE for every phase are in the full guide, which also covers the core work that carries no external load.
One consequence worth naming: in three of the four phases the plan uses a goblet squat, not a barbell back squat. If you enter a barbell max, the calculator will say so rather than convert between the two, because a goblet squat max is not a fixed fraction of a barbell squat max.
The 2026 meta-analysis of strength training in cyclists found significant improvements in cycling efficiency, anaerobic power and performance, and no change in VO2max or the power at VO2max. Strength work makes you better at using the engine you have. It does not build a bigger one.
That is still worth two sessions a week, and it is worth saying out loud, because most pages in this corner of the internet imply the opposite. If you are lifting to add watts at threshold, the evidence does not support the expectation, and knowing that up front is what stops people quitting in week six when the FTP test comes back flat.
What it does buy you: force you can put through the pedals in a sprint or on a steep ramp, resistance to the fatigue that arrives late in a long ride, and a body that handles a season of riding without the usual complaints. None of those show up in an FTP test.
Not your best ever, and not a set from a different variant of the lift. If your last squat session was three sets of eight and the last one was genuinely hard, that is the set to enter.
The calculator already rounds to the nearest 2.5 kg or 5 lb, because a barbell does not do decimals. If your dumbbells jump in bigger steps, take the lower one and add a rep.
The percentage stays the same; the max underneath it moves. In the first two phases it moves quickly, which is the whole point of them.
If the prescribed weight lands two RPE points away from the target on the day, the target wins. RPE is the instruction; the weight is the guess at how to hit it.
Four things, said plainly, because each one is a way a number like this gets misread.
There is no percentile table on this page and that is deliberate. Our app records the RPE and the notes when you finish a strength session, never the load, so we cannot tell you where you rank without inventing it. When we do have the measured data, as with watts per kilo, we publish it.
A squat max says nothing reliable about your row. The bodyweight ratios that circulate for this have no source worth citing, so the calculator works one lift at a time.
Prediction equations are at their best between one and about eight reps. Enter a set of 15 and the range widens, which is why the panel shows a range at all.
A number is not permission. If a movement is new to you, the first weeks are about the pattern, which is exactly what the first phase of the plan is for.
The RPE scale and the meta-analysis are peer reviewed. The two one-rep-max equations are standard prediction formulas from the strength and conditioning field, not clinical trials, and they are listed as such.
Close enough to program with, and not close enough to brag about. Between one and eight reps the two equations on this page usually land within a few percent of each other and within about five percent of a tested max. Past ten reps the spread grows, which is why the panel shows a range instead of a single figure. For picking a working weight, that precision is plenty; the RPE target on the day corrects the rest.
Almost certainly not, if you are a cyclist. A genuine max attempt carries injury risk and a recovery cost that competes with your riding, and it buys you a number you could have estimated from a set you were doing anyway. Powerlifters test because the max is the goal. For you it is only an input.
Because the goal changes. Maximum strength work sits near 83 percent of your max for four reps, because force production needs heavy and brief. The other three phases sit near 65 to 70 percent for ten or twelve reps, because their job is movement quality or holding what you built without stealing from your riding. Same lift, same max, different percentage.
The plan is built for that. Three of the four phases use a goblet squat, a dumbbell Romanian deadlift and lunges, all of which work with whatever load you have, and the first phase is explicitly written for a backpack if that is what is available. The calculator still applies: estimate the max for the variant you actually do, and take the percentage of that.
No, and it will tell you so. Planks, dead bugs and Pallof presses are prescribed by time or by reps, with no external load to calculate. They are in the plan and they matter for transferring power through the bike, but there is no weight for this page to return.
