Linear progression is a training method where you add a small, fixed amount of weight to the bar every session. It stops working when your recovery can no longer keep pace with those jumps, usually after months of consistent training for a beginner. The stall is not a sign you are broken. It is a sign the program has done its job and the next stage needs a different structure.
The hard part is deciding what "stopped working" means, because one bad week and a real structural stall look identical on the day. This guide gives you a way to tell them apart, a clear order of fixes (reset, adjust, or change program), and a map of what intermediate programs do differently. It is written for self-coached lifters, so every rule below is something you can apply from your own training log, without a coach watching.
What is linear progression, and why does it work for beginners?
Linear progression adds a fixed load increase after every successful session, for example 2.5 kg on the bench press or 5 kg on the squat. It works for beginners because an untrained lifter adapts faster than recovery becomes a limit, so almost any consistent, slightly harder session produces a measurable gain.
Three things make this model effective early on. First, a new lifter is responding to an unfamiliar stimulus, so small, regular overload is enough. Second, early strength gains come partly from skill: you get better at the movement itself, which lets the same muscle move more weight. Third, the rule is simple. You either completed the work and add weight, or you did not and you repeat it, which keeps decision fatigue near zero and consistency high.
Research on training status backs up why beginners and trained lifters need different loading. In the Rhea and colleagues meta-analysis of 140 studies and 1,433 effect sizes, a mean intensity of 60% of one-rep max produced the largest gains in untrained people, while 80% worked best in trained people. The same analysis found untrained lifters did best training each muscle group three days a week, and trained lifters two.
The American College of Sports Medicine reaches a similar conclusion for novices. Its position stand on progression models recommends loads in an 8-12 repetition-maximum range and two to three sessions a week for people with no resistance training experience. That is the territory of programs such as Starting Strength and the 5x5 programs.
The same logic explains the expiry date. As you get stronger, each extra kilogram demands more adaptation than your body can deliver in 48 hours, so the gap between what you add and what you can recover from closes. Linear progression does not fail because it was a bad idea. It fails because its core assumption, that you recover faster than you add load, stops being true.
How do you know linear progression has really stopped working?
You know linear progression has stopped working when the same lift fails its prescribed reps on consecutive sessions after you have ruled out sleep, food, technique and attendance. A single missed set is normal variation. Repeated misses at an unchanged load, with a log that shows effort already high, is a structural stall.
Start with the cheap explanations, because they are far more common than "I have outgrown the program". Run through these four before touching the programming:
- Sleep: a short run of poor nights can pull your performance down for the whole week. Check whether the missed session followed one.
- Food: a lifter eating well below maintenance on a program that adds weight every session is asking for a stall.
- Technique: film your top set. If bar path or depth has drifted as the load rose, the bar is heavier than the number suggests.
- Attendance: two missed sessions in a fortnight break the rhythm that linear progression depends on.
If those are in order, use your own data. Your log should show the target reps, the load, and how hard each set felt. A set that took every rep at RPE 9 is a different signal from the same set at RPE 7. RPE (rating of perceived exertion) is a 0-10 effort rating, and RIR (reps in reserve) is the same idea counted in reps left in the tank, so RPE 9 means about one rep left.
| Signal in your log | What it suggests | What to do |
|---|---|---|
| One missed set, effort otherwise normal | Normal variation | Repeat the load next session |
| Same lift misses reps on 2-3 consecutive sessions after a bad night or two | Likely recovery, not programming | Fix sleep and food for a week, then retry |
| Same lift misses reps on consecutive sessions, sleep and food in order | Genuine stall on that lift | Reset that lift (see the next section) |
| Reps hit, but every top set sits at RPE 9 or higher and bar speed is slowing | The next jump is too big | Smaller steps or microloading |
| Two or three lifts stall within the same fortnight | Systemic fatigue, not one lift | A planned deload, then reassess |
| You have already reset once and the same lift stalls again | Structure has run out | Move to an intermediate structure |
Notice that the table does not use training age at all. That is deliberate. "Intermediate" is a gradient, not a switch. The ACSM defines an intermediate lifter loosely, as someone with roughly six months of consistent resistance training, but real lifters stall at different times on different lifts. Squat, deadlift and press often stall at different points, and a lift stalling early does not mean the whole program has to change.
If the stall points to recovery rather than structure, the strength plateau guide walks through diagnosing the cause from your training log in more detail. This guide picks up after that diagnosis, when the answer is that the structure itself needs to change.
Should you reset, change one variable, or change program?
Reset first, adjust one variable second, and change programs third, in that order. A reset (dropping the weight and rebuilding) fixes an overreach cheaply. A single-variable change handles a specific mismatch. A new program is the right answer when the same lift stalls again after a reset, because a reset only buys you another few sessions of the same structure.
The order matters because each step costs more than the last. A reset costs a few weeks of rebuilding. Changing one variable, such as adding a back-off set or dropping a day of volume, costs almost nothing but only helps when the problem was a specific one. Changing programs costs the most: you lose the familiarity of the old structure and have to learn a new set of rules.
Reset. A reset means dropping the stalled lift to a lighter weight and climbing back. Programs differ on how far. GZCLP, for example, tells you to test a new five-rep max after the last stage fails and restart at 85% of it, according to The Fitness Wiki's summary of the program. A reset buys fatigue clearance and a clean run at weights you once handled. It does not change why you stalled, which is why a second stall on the same lift after a reset is a strong signal.
One variable. If you hit the wall on one lift and not others, change only that lift. Swap the step size for smaller jumps, add a back-off set after your top set (the top set and back-off guide covers how), or slow the progression from every session to every second session. This keeps the rest of the program running while the stalled lift catches up.
New program. A simple decision rule is to reset once. This is a working rule, not a research result. If the same lift stalls again, move that lift, or the whole program, to an intermediate structure. Other coaches and forum communities suggest different thresholds, from one reset to three, and the evidence on this exact question is thin. What matters is that you pick a rule before you stall, so you do not drift through five resets because each one felt like it might be the one.
Avoid the opposite failure as well. Constantly hopping to a new program after one rough fortnight resets your progress without ever testing the structure, which is the pattern covered in the program hopping guide. Give a new structure at least one full cycle before judging it.
What do intermediate programs do differently?
Intermediate programs move the unit of progress from the session to the week or the cycle. Instead of adding weight every workout, they separate heavy, light and volume work, stage the rep schemes, or let effort decide the next load. The goal is the same overload, delivered at a pace your recovery can sustain.
The ACSM position stand describes the shift in plain terms. For intermediate to advanced training it recommends "a wider loading range from 1 to 12 RM in a periodized fashion with eventual emphasis on heavy loading (1-6 RM) using 3- to 5-min rest periods between sets." It also advises that when you train at a specific RM load, a "2-10% increase in load be applied when the individual can perform the current workload for one to two repetitions over the desired number."
Put simply, load increases become conditional. You earn the next jump by exceeding the target, rather than receiving it every session. The programs below are different ways of building that condition into a weekly structure.
| Approach | How progress is delivered | Suits you if | Trade-off |
|---|---|---|---|
| Texas Method | Weekly: volume day, light day, then a new five-rep max on the intensity day | You want a clear weekly target and train three days a week | Weekly maxes usually last only a few weeks before the structure needs to change |
| GZCLP | Staged rep schemes per lift; fail a stage and the scheme changes, not the weight | You want a forgiving structure that still has rules | Smaller jumps and a reset rule that is easy to misread |
| Wendler's 5/3/1 | Four-week cycles from a training max, with a deload in every cycle | You want slow, planned progress and built-in deloads | Progress is measured in months, which can feel slow |
| Double progression | Add reps within a range, then add load and drop back to the bottom of the range | Accessory lifts and stalled main lifts that need smaller steps | Needs honest logging of reps and effort |
| Top set plus back-off sets (RPE-based) | One heavy set at a target RPE sets the day's load, lighter sets follow | You are comfortable rating effort accurately | The quality of your RPE judgement caps the quality of the plan |
Texas Method. On IronLedger's Texas Method page, the week is built from Mark Rippetoe's and Glenn Pendlay's published descriptions. Monday is 5 sets of 5 at about 90% of Friday's five-rep max, Wednesday is a light day, and Friday is one set of 5 at a new five-rep max or within 2% of it. Rippetoe places it after roughly five months of novice training. Andy Baker writes that new weekly five-rep maxes usually last four to eight weeks, after which the Friday work moves to heavier sets of one to three rather than the program being dropped.
GZCLP. The Fitness Wiki summarises Cody LeFever's program: the main lift (Tier 1) runs 5x3+, then 6x2+, then 10x1+, moving to the next stage only when you fail the reps at the current one. The secondary lift (Tier 2) steps from 3x10 to 3x8 to 3x6. Notably, the wiki also says AMRAP (as many reps as possible) sets should not go to true failure: "leave 1-2 reps 'in the tank'." It is still linear progression in places, but it turns a stall into a built-in change of scheme instead of a dead end.
5/3/1 and the training max. Wendler's approach plans load from a training max, a deliberately conservative number below your true max, across a four-week cycle. The built-in guided 5/3/1 in IronLedger uses a default training max of 90% of a tested or estimated one-rep max, which you can adjust, and shows the percentage behind each load. The method suits lifters who are tired of guessing, since progress is planned rather than negotiated session by session. Lifters coming off a stall often pair it with the training max calculator.
None of these is the single right answer. What they share is conditional, slower progress. If you keep that principle and pick the structure that suits how you like to train, the specific choice matters less than sticking with it for a full cycle.
How do you use RPE and RIR when progress comes weekly instead of every session?
You use RPE and RIR to decide whether the next increase is earned. If you hit the target reps at a moderate effort (RPE 7 or below, three or more reps in reserve), add weight. If you hit them at RPE 9 or higher, or miss reps, hold the load. The rating turns a vague "that felt heavy" into a decision rule you can apply in the same way every week.
The scale has a research basis. In Zourdos and colleagues' 2016 study, 29 squatters (15 experienced, with an average training age of 5.2 years, and 14 novices at 0.4 years) rated single reps at set percentages of their one-rep max. RPE 10 meant zero reps in reserve and RPE 9 meant one. Bar speed and RPE showed a strong inverse relationship in both groups (r = -0.88 for experienced lifters, r = -0.77 for novices), and the experienced squatters rated a true one-rep max higher than novices did. The authors concluded the RIR-based scale is a practical way to regulate daily training load.
There is an important caveat for self-coached lifters. Later work from the same group found that rep-in-reserve estimates are less accurate when you are far from failure and in higher-rep sets, and more accurate close to failure. In practice, that means RPE is most trustworthy on heavy sets of low reps. Rate your top set of three or five carefully, and treat the RPE of a set of fifteen as a rough guide only. You can sanity-check any logged set with the RPE calculator, which converts load, reps and effort into an estimated max.
Two RPE-based progression rules cover most intermediate situations:
- Top set to top set. Work up to one heavy set at a target RPE (for example, a set of five at RPE 8). Next week, aim for the same reps at the same RPE with slightly more weight. If the weight went up and the RPE stayed the same, you are progressing. If RPE crept up for the same weight, you are not.
- Rep range, then load. Pick a rep range such as 4-6. When every set reaches the top of the range at RPE 8 or lower, add load and return to the bottom. This is double progression with an effort check, which stops a lifter from grinding out the top of a range on a bad day.
Autoregulation, adjusting the plan to how you perform on the day, also has some direct trial evidence. In Mann and colleagues' 2010 study, 23 Division I football players with an average of 2.65 training years followed either an autoregulatory program or a linear periodised program for six weeks. The autoregulated group improved bench press one-rep max, estimated squat one-rep max and repetitions at 225 lb more than the linear group. Treat that as an encouraging signal, not a verdict: the sample was small, the trial was six weeks long, and the athletes were trained collegiate players under supervision. Whether the same gap appears for a self-coached lifter over a year is unproven.
For a longer treatment of the method, see the guides to autoregulated training and RPE versus percentage-based training.
How much volume and frequency does an intermediate lifter need?
Most intermediate lifters need slightly more weekly work than beginners, spread over three to four sessions, but not dramatically more. The ACSM recommends two to three sessions a week for novices, three to four for intermediates and four to five for advanced lifters, and the research on sets suggests more volume helps muscle size up to the limits of what you can recover from.
Two findings are worth holding together, because they appear to pull in different directions. The Rhea meta-analysis found that trained lifters did best training each muscle group about two days a week, with four sets per muscle group per session producing maximal gains in both trained and untrained groups. The ACSM's three to four sessions a week for intermediates is a count of total sessions, not sessions per muscle group, so a three-day full-body week and a four-day upper/lower week can both fit.
For muscle size specifically, the Schoenfeld, Ogborn and Krieger meta-analysis pooled 34 treatment groups from 15 studies and found each additional weekly set was associated with a 0.37% larger gain in muscle size, a graded relationship rather than a threshold. That result is about hypertrophy, not maximal strength, so it supports adding volume when your goal includes size, but it does not tell you how many sets will raise your squat.
| Variable | Novice (ACSM 2009) | Intermediate (ACSM 2009) | What to log |
|---|---|---|---|
| Sessions per week | 2-3 | 3-4 | Sessions completed, not planned |
| Load range | 8-12RM | 1-12RM, periodised, heavy 1-6RM emphasis | Load and reps on every working set |
| Rest between heavy sets | Not specified in the abstract | 3-5 minutes | Rest time on main lifts |
| Load increase rule | Add each session | 2-10% when you beat the target by 1-2 reps | Reps beyond target, and RPE |
| Progress delivered | Every session | Weekly or per cycle | Weekly top set and its RPE |
Use these as ranges, not prescriptions. The ACSM table summarises group-level evidence, and individual responses to volume vary with sleep, food, stress and how long you have trained. A sensible way to apply it is to start at the low end of the volume range for your new program, hold it for a few weeks, and add sets only when your log shows you finishing sessions with reps in reserve and your top sets still moving. For help setting a starting number, read the guides on sets per muscle group and training frequency per muscle group.
Rest is part of the same decision. Heavier work with 3-5 minutes between sets takes longer per session than the novice routine, so build the extra time into your week. A program that quietly needs 90 minutes a session will start losing attendance, and attendance was the first cheap explanation to rule out when you stalled.
How do you track the transition in IronLedger?
IronLedger gives you the record a transition decision depends on: load, reps and RPE or RIR for every set, personal records and estimated one-rep max from what you actually lifted, and a full history with no cutoff. It is a free strength training app and logger, so you can run the whole decision from your own data without a coach reviewing it.
The product page lists what the app records per set: weight, reps, RPE in half steps (0-10) or RIR (0-5), and a free-text note, with a rest timer that keeps counting when your phone is locked. The progress analytics view adds personal records, estimated one-rep max (calculated with the Epley formula) and per-lift trends and tonnage. Those are the same signals the table earlier in this guide asks you to read: repeated misses, rising RPE for the same load, and flat numbers across weeks. The estimated one-rep max guide explains how the estimate is calculated and when to trust it.
Here is how the pieces map to the decision rules:
- Spotting the stall: per-lift history shows whether the last three sessions of a lift hit target reps, and the RPE you logged shows whether it was close.
- Resetting cleanly: a note on the set records why you reset, so you can see later whether the reset worked.
- Switching program: IronLedger has ten program templates and a custom multi-week builder with deloads, so a Texas Method style week or a 5/3/1 cycle can live in the same history as your old linear block. You can also bring a spreadsheet in as a file: the free GZCLP, nSuns and Texas Method pages generate import files from your own numbers.
- Progressing by effort: on its five built-in programs, the app's load recommendation adds 2.5 kg or lb after a session where every prescribed rep was hit at RPE 7 or lower (or RIR 3 or more, or no effort logged), and holds after missed reps or a hard session at RPE 9 or higher. The rule is deterministic, and how IronLedger decides your next weight shows the reasoning behind each number.
Two limits are worth knowing before you plan around the app. Live load recommendations run only on the five built-in programs, so an imported file keeps the loads it was written with and you adjust missed weeks yourself. And Health Connect, Strava and Oura appear as "coming soon", not as available integrations, while Apple Health syncs completed strength workouts on iPhone. The app is available on iPhone now, with Android in closed testing, per the download page.
The main thing the log gives you is a defensible answer to the question this guide started with. Instead of asking "has linear progression stopped working?" on a bad day, you can open three weeks of sets and read it.
Frequently asked questions
Stop when the same lift fails its reps on consecutive sessions after you have ruled out sleep, food, technique and missed sessions. A single missed set is normal. Check your logged RPE too: top sets all at RPE 9 or higher suggest the jumps are too big.
Sources and references
- American College of Sports Medicine. (2009). Progression models in resistance training for healthy adults. Medicine & Science in Sports & Exercise, 41(3) — novices use 8-12RM loads and 2-3 sessions a week; intermediates use a periodised 1-12RM range with eventual emphasis on 1-6RM loading, 3-5 minute rests, and 3-4 sessions a week; apply a 2-10% load increase when the current load is exceeded by one to two repetitions (2009).
- Rhea, M.R., Alvar, B.A., Burkett, L.N., & Ball, S.D. (2003). A meta-analysis to determine the dose response for strength development. Medicine & Science in Sports & Exercise, 35(3) — 140 studies and 1,433 effect sizes; about 60% of one-rep max was most effective for untrained lifters and 80% for trained lifters, with trained lifters training each muscle group two days a week and four sets per muscle group best for both groups (2003).
- Zourdos, M.C. et al. (2016). Novel resistance training-specific rating of perceived exertion scale measuring repetitions in reserve. Journal of Strength and Conditioning Research, 30(1) — 29 squatters; strong inverse relationship between bar velocity and RPE (r = -0.88 experienced, r = -0.77 novice); the RIR-based scale is a practical method to regulate daily training load (2016).
- Mann, J.B., Thyfault, J.P., Ivey, P.A., & Sayers, S.P. (2010). The effect of autoregulatory progressive resistance exercise vs. linear periodization on strength improvement in college athletes. Journal of Strength and Conditioning Research, 24(7) — 23 Division I football players over six weeks; the autoregulatory group improved bench press and estimated squat one-rep max more than the linear periodisation group (2010).
- Schoenfeld, B.J., Ogborn, D., & Krieger, J.W. (2017). Dose-response relationship between weekly resistance training volume and increases in muscle mass. Journal of Sports Sciences, 35(11) — 34 treatment groups from 15 studies; each additional weekly set was associated with a 0.37% increase in the percentage gain in muscle size (2017).
- The Fitness Wiki. GZCLP routine summary, based on Cody LeFever's program — Tier 1 stages of 5x3+, 6x2+ and 10x1+, Tier 2 stages of 3x10, 3x8 and 3x6, a reset to 85% of a new five-rep max, and AMRAP sets taken with one to two reps in reserve (reference summary, accessed 2026).
Keep learning
