Percentage of 1RM for strength: NSCA + ACSM 2026 loading bands

Strength training percentages live in a narrow window. The NSCA traditional band runs 80–95% of 1RM; the ACSM 2026 Position Stand on Resistance Training reframes the same target as a ≥80% thresholdwith intent and sufficient effort. Both framings put strength work above 80% and pair it with low rep counts (1–6), low Reps in Reserve (1–3), and long rest. The load table on this page is computed off a 100 kg 1RM through the same percentage-of-1RM calculator you'd use for your own number.

The two framings overlap — but for different reasons. NSCA periodizes within an 80–95% window because most published programs cycle the working percentage week to week. ACSM 2026 collapses that window into a single threshold because the meta-analytic evidence (137 systematic reviews, ~30,000 participants) is that anything ≥80% with intent drives strength adaptation. The methodology page carries the full dual-framing table; this article walks through what to do with the strength row.

The strength row, NSCA + ACSM 2026 side-by-side

The percentages engine returns the strength prescription as one row with two intensity framings, one set/rep prescription, and a single RIR target. Pulled directly from lib/strength/percentages.ts:

The verdict embedded in the engine is to default the load table to NSCA 80–95% for in-table highlighting and treat the ACSM 2026 ≥80% threshold as the wider permission slip. Most lifters grew up programming inside the NSCA window — 5×5 at 85%, 4×4 at 87.5%, heavy singles at 92.5% — and the strength outcome is good. ACSM 2026 doesn't invalidate any of that; it says the lower edge (80%) is where the adaptation starts and there's no penalty for spending most of the work above 90% if the recovery supports it.

The load table at 100 kg

A 1RM of 100 kg run through the percentages engine returns the load table below. Rows in the NSCA strength band (80–95%) are marked; rows below 80% are sub-strength; the 100% row is the input itself. The engine rounds kg loads to the nearest 0.5 kg (matching the smallest competition micro-plate, 0.25 kg/side), rounds lb loads to the nearest pound, and the load table caps at 100% by default — test attempts at 102.5% / 105% require explicit max-test mode.

The numbers look clean because 100 kg is the textbook example — plug a real 1RM in and the rounding starts to bite. A 137 kg bench-press 1RM produces 109.5 kg at 80% and 116.5 kg at 85%; a 137 kg squat at 92.5% rounds to 126.5 kg. If your gym only stocks 2.5 kg as the smallest plate, the 5% rows you can actually load are sparser than this table suggests — many of them snap to the nearest 5 kg increment once you halve and round per side. The 0.5 kg rounding posture in the engine is for programming, not loadouts; the plate calculator handles plate-level rounding separately.

What ACSM 2026 actually changed for strength

The ACSM 2026 Position Stand on Resistance Training (Currier et al., Med Sci Sports Exerc 58(4):851–872) is the first major ACSM update in this area since 2009. Its strength prescription is one sentence:

“Voluntary strength was enhanced by lifting heavier loads (≥80% one-repetition maximum), through a complete range of motion, for 2–3 sets, at the beginning of training sessions, and ≥2 sessions/wk.”

That's a different shape from a periodized 80–95% band. NSCA publishes 80–95% because real-world programs cycle through it — a 5×5 block at 80–85%, a 4×4 block at 85–90%, heavy singles in the 90–95% range as a peak. ACSM 2026 says the band itself isn't the variable that matters: ≥80% with intent and effort drives the adaptation. Where the two views fight is at the upper end. NSCA-style programming routinely puts heavy singles in the 92.5–95% range as part of normal training; ACSM 2026 has nothing against that, but its summary doesn't require it. If a beginner spends a year at 80–85% with sufficient effort and never touches a 90% set, ACSM 2026 is fine with that outcome.

The take embedded in the engine: NSCA 80–95% is the default load table because it matches what most written programs do, and ACSM 2026 is the wider claim that says the upper portion of that band isn't mandatory. Read the strength row as “at least 80%, mostly inside 80–95%, with the 95%+ work earned through accumulated training, not assumed.”

Sets, reps, and rest the engine recommends

The strength prescription returned by the engine is 3–6 sets × 1–6 reps with 2–5 minutes rest. That's a wider envelope than “5×5” or “3×3” — the range covers everything from a high-volume strength block (6×4 at 80%) to a peaking-adjacent week (3×2 at 92.5%). The two pieces worth highlighting are the rep ceiling and the rest floor.

Reps cap at 6 because once the working set runs past six reps with 1–3 RIR you're no longer at strength loads — the math forces you below 80% to finish the set, which slides you into hypertrophy territory. Rest starts at 2 minutes and stretches to 5 because incomplete recovery between heavy sets converts a strength session into a fatigue-management session, and the literature on heavy-load rest intervals (NSCA Essentials chapter on resistance-training program design) consistently recommends 2 minutes minimum at ≥80% loads, with 3–5 minutes for the heaviest sets. ACSM 2026 reads as “2–3 sets minimum” on the bottom end; the engine's 3–6 range stays inside the periodized strength block most coaches actually program.

The opinion: 5 sets × 5 reps at 80–85% is the most-bang-for-buck inside this row for an intermediate lifter — it accumulates enough working volume for hypertrophy carryover while staying mechanically heavy enough that the strength stimulus is doing real work. 3×3 at 90% is the answer when the goal is neural strength without the fatigue tail. Anything above 5×5 with 1 RIR on a 5-rep set is a programming choice with a recovery cost; a 6×6 at 82.5% is a hard week, not a weekly default.

RIR as a supplementary intensity quantifier

Reps in Reserve — how many additional reps the lifter could complete with good form on the working set — is the intensity dial ACSM 2026 formally added alongside %1RM. The engine returns RIR 1–3 for strength, which tracks the near-maximal-but-not-grinding range most heavy strength sets actually live in.

%1RM is the better planning tool — it's a fixed number, easy to write into a program, and tied directly to the input 1RM. RIR is the better in-the-moment tool, because the same 85% load is a different effort on different days. A working set planned at 85% × 5 might come in at 1 RIR on a fresh session and at 3 RIR on a deload-tail Wednesday. The engine's 1–3 range covers that variance honestly: 1 RIR is “could have done one more clean, but barely”; 3 RIR is “heavy, controlled, never in doubt.”

When RIR-driving beats %1RM-driving: deload weeks (cap RIR at 4–5 regardless of the percentage), competition prep tapers (drop RIR to 1 on top sets while %1RM stays constant), and any week where the lifter knows their estimated 1RM is stale. When %1RM-driving wins: writing a 12-week block in advance, comparing programs, and working with a lifter whose recent 1RM is fresh and reliable.

What the strength band does NOT do

The strength row tells you which loads to lift and how many times to lift them. It says nothing about technique drift, equipment differences, fatigue state, or coaching context. The full list of what the calculators don't model lives on the methodology page. The two that bite hardest in strength work specifically are fatigue state and bar-path drift on grindy 5-rep sets — both can turn a planned 85% × 5 into a 3 RIR easy day or a 0 RIR miss with identical numbers in the spreadsheet.

The closing take: percentage-of-1RM is a planning tool, not a prescription. The strength row gives you the window (≥80%, mostly 80–95%), the rep ceiling (six), and the RIR floor (one). What you do inside that window — peaking, accumulation, deload timing — is coaching, not math. The engine surfaces the math; the rest stays on the lifter.

Common questions

What's the best %1RM for strength training?

≥80% of 1RM per the ACSM 2026 Position Stand. The NSCA traditional 80–95% band — the field default for ~30 years — sits inside that threshold and remains the most-programmed range. Both produce strength gains; ACSM 2026 is broader, not contradictory. The StrengthMath load table highlights 80–95% as in-band rows because that's the periodized window most written programs operate in.

Should I use sets of 3 or sets of 5 for strength?

The percentages engine returns 3–6 sets × 1–6 reps with 2–5 minutes of rest for strength. Specific within that range is goal-dependent: 5×5 at 80–85% builds a wider rep base and tolerates more weekly frequency; 3–5 sets of 1–3 reps at 90%+ is heavier and lower-volume per session, useful for training-max prep or peaking. RIR 1–3 ties both ends together — the working sets should be near maximal effort, not grinders.

Is RIR or %1RM the better way to gauge intensity?

Both. RIR captures the actual stimulus on a given day — the same 85% of last week's 1RM is a 1-RIR set when you're rested and a 3-RIR set after a hard week. %1RM is fixed against an estimated 1RM that may drift week to week. ACSM 2026 explicitly endorses RIR alongside %1RM as a supplementary intensity quantifier. Use %1RM to plan the session; use RIR to read the bar in front of you.

How does this relate to the 1RM I plug in?

Every percentage in the load table is computed off your input 1RM, so the upstream question is which 1RM number to trust. The companion guide on the four common 1RM prediction formulas — Epley, Brzycki, Lombardi, O'Conner — covers per-lift bias and the LeSuer 1997 validation; deadlift in particular is under-predicted by 9–14% across all four formulas. If your input 1RM is wrong, every row in the strength load table is wrong by the same factor.

Where to next

Strength sits inside a larger force-velocity picture. The companion guide on percentage of 1RM for power covers the ACSM 2026 widening from 50–75% to 30–70% on the power row — different mechanism (rate of force development), different load range, but the same dual-framing pattern. If your input 1RM itself is in question, the best 1RM formula guide covers the per-lift accuracy of Epley, Brzycki, Lombardi, and O'Conner against the LeSuer 1997 validation data — useful before you trust any load row in this article. For the live calculator, the percentage-of-1RM calculator carries both NSCA and ACSM 2026 framings on every goal — strength, power, hypertrophy, speed, deload, and peaking.