I use the RPE scale for runners as a 1–10 control system: 2 is recovery, 3–4 is easy, 5–6 is steady, 7 is controlled hard, 8–9 is hard to very hard, and 10 is maximal.
The number matters less than the anchors behind it. I pair RPE with speech, breathing, duration, mechanics, and the purpose of the session. That turns “run by feel” from vague advice into a skill I can repeat when pace, weather, terrain, or heart rate changes.
RPE is not a toughness score. A higher number is not a better run. The correct number is the effort that completes today’s assignment while preserving what comes next.
My complete 1–10 running RPE scale
| RPE | What it feels like | Speech | Typical use |
|---|---|---|---|
| 1 | Barely working | Unrestricted | Standing or gentle walking |
| 2 | Very easy | Effortless conversation | Warm-up, cooldown, recovery jog |
| 3 | Easy | Full relaxed sentences | Most easy running |
| 4 | Comfortable but purposeful | Full sentences with deeper breathing | Long aerobic running |
| 5 | Steady | Shorter sentences | Moderate steady work |
| 6 | Firm and sustainable | Brief sentences | Marathon-like effort for trained runners |
| 7 | Controlled hard | Short phrases | Threshold intervals or tempo |
| 8 | Hard | A few words | 5K–10K-style repetitions |
| 9 | Very hard | Single words | Short severe intervals or late racing |
| 10 | Maximal | No speech | Final sprint or true test—not routine training |
These are practical anchors, not laboratory zones. A 20-minute effort and a two-hour effort cannot be interpreted without duration. Heat, hills, training history, and fatigue also change what speed corresponds to each number.
What I am actually rating
My question is: “How hard am I working to sustain this running task right now?”
I do not ask how much I enjoy it, how motivated I feel, or how impressive the pace looks. I also keep effort separate from pain, soreness, and local muscular discomfort.
A sore calf can feel uncomfortable while aerobic effort remains low. That does not make the run RPE 8; it makes the symptom a separate decision. Pain that changes gait is not a high training score—it is a reason to stop and assess.
I often record three short notes when calibration matters:
- whole-body effort;
- breathing demand;
- leg or local discomfort.
That separation tells me whether the limiter is cardiovascular, muscular, environmental, or potentially medical.
My five-anchor shortcut
I do not expect ten perfectly distinct sensations. I calibrate five anchors and let the space between them hold normal variation.
| Anchor | My cue | What must remain afterward |
|---|---|---|
| RPE 2 | Recovery; almost deliberately slow | No added fatigue |
| RPE 3–4 | Easy; conversational and relaxed | Capacity for the next planned session |
| RPE 5–6 | Steady; clearly working but patient | Stable mechanics and no desperate finish |
| RPE 7 | Controlled hard; short phrases | Repeatability and one more controlled repetition |
| RPE 8–9 | Hard to very hard | Quality form for the prescribed short duration |
I reserve 10 for truly maximal work. If RPE 10 appears every week, the scale has become a reward system for suffering rather than a tool for training.
My broader guide to training by effort instead of pace shows how these anchors replace rigid pace targets without removing structure.
The three cues I use before the number
Speech
Full relaxed sentences support an easy rating. Shorter sentences fit steady running. Short phrases belong near controlled threshold, and isolated words signal high intensity.
I use a familiar sentence rather than casual conversation with a talkative partner. Personality should not set the zone.
Breathing
I notice rhythm and control, not just loudness. Deep breathing can remain organized. Ragged breathing that does not settle usually means the effort is higher than intended.
Time remaining
I ask, “Could I continue this effort for the duration the session requires?” RPE 7 for three minutes is not equivalent to RPE 7 for forty minutes.
The number comes after those observations. That order reduces the temptation to choose a number first and defend it afterward.
My two-week calibration block
I use two weeks to make the scale personal without turning every run into a test.
Week 1: establish the anchors
On two easy runs, I hide pace and check speech, shoulders, hands, breathing, and whole-body effort at 10-minute intervals. I reveal pace and heart rate only afterward.
For the controlled-hard anchor, I run 3 × 8 minutes at what I believe is RPE 7 with two minutes of easy jogging. The repetitions should be even enough that the last is demanding but not desperate.
If the first repetition is fastest and speech disappears immediately, my “seven” is really closer to eight or nine.
Week 2: test repeatability
I repeat one familiar easy route under reasonably similar conditions. On another day, I use a progression from RPE 3 toward 6 without crossing into threshold.
After each session, I compare the feeling with pace, heart rate, weather, and the next-day response. I revise the written anchors—not the memory of what I hoped the run was.
The goal is consistency, not perfect precision.
My RPE cards for common running sessions
| Session | Working RPE | My execution rule | Sign I have gone too hard |
|---|---|---|---|
| Recovery run | 2–3 | Short, optional, mechanically normal | It creates fatigue instead of reducing stiffness |
| Easy run | 3–4 | Full conversation; pace floats | Speech shortens or I defend pace |
| Long easy run | 3–4, sometimes 4 late | Start with margin; fuel when appropriate | Unplanned rise to 6 with declining form |
| Steady run | 5–6 | Purposeful, patient, below threshold | It becomes fragmented-speech running |
| Threshold | 7 | Controlled hard and repeatable | The first repetition feels like a race |
| Long intervals | 7–8 | Preserve mechanics across repetitions | Pace collapses while effort spikes |
| Short intervals | 8–9 | High quality for limited duration | Every repetition becomes a sprint |
| Race | Builds with distance | Match early restraint to event length | RPE 9 arrives before the race has developed |
For ordinary aerobic days, I follow the conversational guardrails in easy runs explained. For moderate work, steady runs explained keeps RPE 5–6 from quietly taking over the week.
Threshold is the most commonly inflated label. My threshold-run system treats RPE 7 as controlled and repeatable, never as a disguised time trial.
How I combine RPE, talk test, heart rate, and pace
I use a hierarchy rather than averaging conflicting numbers:
- Session purpose.
- RPE and speech.
- Heart-rate trend.
- Pace or power output.
When all four agree, the decision is easy. When they conflict, I investigate.
If pace is slow but RPE and breathing are high on an easy day, I do not chase pace. I review heat, hills, sleep, illness, fueling, and recent training.
If heart rate jumps while breathing stays calm, I suspect the sensor. My guide to wrist versus chest heart-rate monitors helps me identify cadence lock, poor contact, or chest-strap issues.
If heart rate climbs gradually with stable pace, I use the diagnostic process in heart rate drift while running rather than assuming fitness or hydration from one graph.
Conditions change pace—not the assignment
RPE becomes most valuable when the environment makes pace unreliable.
| Condition | What changes | My RPE decision |
|---|---|---|
| Heat and humidity | Cooling demand and heart rate rise | Preserve target RPE; accept slower pace; shorten if effort keeps rising |
| Uphill | Speed falls while metabolic demand rises | Hold the intended effort, not flat-road pace |
| Downhill | Breathing eases but impact and braking load remain | Keep mechanics controlled; low RPE does not erase tissue stress |
| Headwind | Pace becomes expensive | Hold posture and effort; stop fighting the watch |
| Altitude | Familiar pace requires more ventilation | Reduce pace and allow acclimatization |
| Trail terrain | Footing and grade change continuously | Rate the whole task; use time and effort more than road splits |
For technical or hilly terrain, my trail effort versus road pace separates cardiovascular demand from the mechanical cost that RPE alone can miss.
For structured climbs, I use the repetition and form rules in hill repeats for runners so a target RPE does not become permission to sprint the first ascent.
RPE drift is information, not failure
At a constant pace, effort usually rises with duration. Body temperature, fluid loss, carbohydrate availability, fatigue, wind, and terrain can contribute.
An easy long run moving from RPE 3 to 4 may still match its purpose. A rise from 3 to 6 with deteriorating form means the session has changed even if pace has not.
I record RPE at stages—early, middle, and late—rather than forcing one number to describe the entire run.
If RPE rises unexpectedly, I slow first. Then I check conditions, fueling, hydration, mechanics, and recent load. I use the duration rules in long runs explained to decide whether continuing adds useful training or merely extends fatigue.
Interval RPE and session RPE answer different questions
During a workout, repetition RPE tells me whether each work bout has the intended intensity. Session RPE tells me the total internal cost of warm-up, work, recoveries, and cooldown.
Five repetitions may reach RPE 8 while the complete session rates 6 because recoveries are generous. A long progression may never exceed 7 but still produce a high session rating through duration.
I record session RPE about 15–30 minutes after finishing and keep that timing consistent. At the finish line, the final repetition dominates. Many hours later, memory and unrelated fatigue distort the answer.
Both numbers matter: repetition RPE protects execution; session RPE helps me understand the dose.
The simple session-load calculation I use
I estimate internal load as:
session RPE × duration in minutes
A 40-minute run rated 3 produces 120 arbitrary units. A 60-minute run rated 6 produces 360.
I compare those values only within my own training and only when rating instructions stay consistent. The score is not exact physiology, and it cannot make two athletes comparable.
It also cannot see everything. A technical trail run and a flat road run may share the same product while stressing tissues differently. I still record terrain, elevation, workout type, and symptoms.
A worked week through my RPE lens
| Day | Planned session | Momentary RPE | Session RPE | My interpretation |
|---|---|---|---|---|
| Monday | Rest | — | — | Absorb the weekend |
| Tuesday | 3 × 8 min threshold | 7 during work | 6 | Correct if repetitions remain even |
| Wednesday | 45 min easy | 3 | 3 | Recovery supports the week |
| Thursday | Easy + strides | 7 briefly in strides | 3–4 | Short peaks do not redefine the run |
| Friday | Rest or short strength | Varies | Recorded separately | Do not hide lifting cost |
| Saturday | 35 min easy | 3 | 3 | Finish with margin |
| Sunday | 90 min long run | 3 early, 4 late | 4 | Normal controlled drift |
This distinction stops one fast minute from misclassifying an easy run. It also exposes days labeled “easy” that spend most of their duration at 5–6.
I place real recovery using my rest-day scheduling rules. A low-RPE run is still training, and a hard gym session does not become recovery because it contains no miles.
My high-RPE and low-RPE decision rules
When RPE is higher than expected, I check five things: conditions, opening pace, sleep and life stress, fuel and fluids, and signs of illness or injury.
One unusual run may be noise. Several sessions with rising RPE, weak pace, poor sleep, heavy legs, and declining motivation justify a load reduction. My cutback-week framework removes the actual stressor instead of simply changing the label.
Unexpectedly low RPE can mean improved fitness, favorable conditions, or fresh legs. It does not automatically authorize extra repetitions. I complete the assignment and look for repeatability.
Low heart rate with high RPE and poor pace is not a medal. It may reflect fatigue, illness, low fuel availability, medication effects, or bad data.
How I coach common calibration errors
Rating discomfort instead of effort
I record local discomfort separately. When delayed soreness is the question, my DOMS decision guide distinguishes ordinary muscle soreness from patterns that need a different response.
Rating only the final minute
I use momentary RPE for a repetition and session RPE for the whole workout. One hard finish should not rewrite forty controlled minutes.
Changing definitions by mood
I keep written anchors. “Easy” cannot mean full conversation on Monday and fragmented speech on Friday.
Using decimals
RPE 6.7 suggests accuracy the sensation cannot deliver. I use whole numbers and describe the trend when needed.
Turning every hard session into 10
Most useful training remains below maximal. RPE 10 has poor repeatability and high cost; I reserve it for rare, appropriate tests or a final race sprint.
RPE on race day
Race RPE depends on distance. A marathon may begin near 5–6 and climb gradually. A 5K reaches high ratings much earlier.
I use the first third to confirm restraint, the middle to protect rhythm, and the final third to decide whether the whole system can increase—not just whether motivation is available.
Adrenaline can make early pace feel artificially easy. I still respect the intended opening effort and check breathing, mechanics, and course conditions.
Reaching RPE 9 too early is a pacing warning. RPE 10 belongs, if anywhere, in the closing sprint when no further work must be preserved.
My medical and pain boundary
RPE measures perceived exertion; it does not diagnose symptoms.
I stop for chest pain or pressure, faintness, confusion, severe unusual shortness of breath, loss of coordination, sharp escalating pain, or a gait change. Urgent symptoms require urgent help, not a revised training number.
Medication, illness, altitude, pregnancy, and cardiovascular or metabolic conditions can change exercise response. Individual medical guidance outranks a generic scale.
The purpose of RPE is better self-regulation. It is never permission to run through a warning sign.
Frequently asked questions
What RPE should easy runs be?
I usually keep easy runs at RPE 3–4 with full conversational speech. Recovery jogging may sit at 2–3. Pace is allowed to change with weather, terrain, and fatigue.
What RPE is threshold running?
I use about RPE 7: controlled hard, short phrases, and repeatable. If the first repetition feels like a time trial or mechanics collapse, the effort is too high.
Is the RPE scale accurate?
It is useful when the question, anchors, and timing remain consistent. It is subjective, so I calibrate it with speech, breathing, pace, heart rate, conditions, and repeated experience.
Should I use a 1–10 or 6–20 RPE scale?
Either can work. I use 1–10 because it is intuitive for most runners. I do not mix scales in the same training log because that destroys trend consistency.
When should I record session RPE?
I usually record it 15–30 minutes after the cooldown and keep the timing consistent. I record repetition RPE separately during structured workouts.
Why does the same pace feel harder on some days?
Heat, hills, wind, sleep, accumulated load, fueling, illness, and stress can raise internal cost. RPE reveals that change so I can preserve the session’s purpose instead of defending pace.
Can RPE replace heart rate and pace?
It can guide training when devices or pace targets are unreliable, but I prefer triangulation. Session purpose and RPE lead; speech, heart-rate trend, and pace help explain the result.
Should a race reach RPE 10?
Only perhaps in the final sprint. Most of the race should remain below maximal effort, with RPE rising according to distance, pacing, terrain, and conditions.
My bottom line
The RPE scale works when every number has a stable job.
I anchor effort to speech, breathing, duration, mechanics, and session purpose. I separate exertion from pain, record momentary and session RPE differently, and let pace change when conditions change.
The skill is not predicting the “correct” number before every run. It is noticing when the experience no longer matches the assignment—and making the next decision early enough to protect the week.
