Rear Delt Exercises: Best Moves for Your Rear Delts
The best rear delt exercises are the reverse pec deck, the cable rear-delt fly and the head-on-bench dumbbell rear-delt raise, because each one isolates horizontal abduction, the rear delt's main job. Add face pulls or wide-elbow rows when you want heavier loading. Start with 6 to 8 hard sets a week on top of your rows, mostly for 12 to 20 reps.
In this guide
What the rear delts do
Your rear delts, or posterior deltoids, are the back third of the deltoid, so they belong to the shoulder. Most rear delt exercises train one or both of their two main jobs. Horizontal abduction moves your upper arm backward at about shoulder height. Shoulder extension brings the arm down and back from in front of you.
The rear delts also help rotate the arm outward. In rows and pulls they work beside the mid traps, rhomboids and lats, which is why this guide sits in the back exercise hub as well as the shoulder exercise hub. For the wider picture, see what the research says about the best back exercises.
Horizontal abduction is the job the lats share least. A row with tucked elbows lets the lats do much of the pulling, while a reverse fly keeps the arm out to the side, where the rear delt does much of the work alongside the mid traps, rhomboids and rotator cuff. That is why flyes lead the list below.
The rear delt exercises worth your sets
| Rank | Exercise | Why it earns the spot | Best for |
|---|---|---|---|
| 1 | Reverse pec deck | Fixed arc, chest support, small load steps | Most lifters in a gym |
| 2 | Cable rear-delt fly | Steady cable tension, adjustable height | Gyms with a cable station |
| 3 | Head-on-bench dumbbell rear-delt raise | Supported reverse fly with only dumbbells | Home and busy gyms |
| 4 | Face pull | Adds outward rotation and heavier loads | Pull days; modelling data only |
| 5 | Wide-elbow rear-delt row | Heavy horizontal pull with flared elbows | Rear delts and upper back together |
| 6 | Band pull-apart | Portable, light, easy to repeat | Warm-ups, home and travel |
The top three are all reverse flyes. They isolate horizontal abduction, keep your torso still, and let you push sets close to failure before your grip or lower back gives out. They differ mainly in resistance profile: the machine guides the arc, the cable keeps tension at the start where dumbbells go slack, and the dumbbell version needs only a bench.
Face pulls and wide-elbow rows rank lower because the traps, rhomboids and rotator cuff share the work, so the rear delts rarely end the set. Cable rope rear-delt rows fill the same slot as the barbell version. Band pull-aparts come last: a band is hardest at the end, so the rear delt gets little load where it is stretched.
What activation studies measure, and what they miss
Most rear delt research measures EMG, the electrical activity of a muscle during an exercise, often reported as a percentage of a maximal contraction (% MVIC). A current-opinion review by Vigotsky and colleagues concluded that acute surface EMG amplitude is not a validated predictor of muscle growth. Read the numbers below as clues about which muscles work, not as growth rankings.
- A fine-wire EMG study in 10 healthy subjects recorded the highest posterior deltoid activity of seven exercises (88% MVIC) in prone horizontal abduction with the arm at 100° and fully rotated outward, which is a reverse-fly position. That position also drove the middle deltoid hard (87% MVIC).
- A systematic review of 33 EMG studies ranked pull-ups and inverted rows among the highest posterior-deltoid activators. It averaged values across mixed methods rather than pooling them statistically.
- An EMG study in 13 trained men measured very low posterior deltoid activity in the bench press (3.5%) and dumbbell fly (2.5%), more in the shoulder press (11.4% MVIC) and the lateral raise (24% MVIC). It tested no rows or reverse flyes.
So a push-heavy program leaves the rear delts with little work. Reverse flyes and pulls both produce high signal, but EMG cannot tell you which builds more muscle over months.
Do rows already train your rear delts?
Yes, partly. Every row extends the shoulder, and the rear delt assists that motion. Elbow position changes the share: tucked elbows hand more of the pull to the lats, while flared elbows turn the row toward horizontal abduction.
A computer-modelling study in 10 men estimated posterior deltoid force at about 59 N for a bent-over dumbbell lateral raise and 58 N for a standing face pull, both higher than about 48 N for a bent-over dumbbell row. Those forces came from a model, not direct measurement.
A crossover EMG study in 14 trained men recorded broadly similar back-muscle EMG in seated rows with the shoulder blades fixed or free, with modestly higher concentric rear-delt signal when they were fixed (effect size 0.66). An EMG study in 16 trained men compared full, top-half and bottom-half prone barbell rows and saw no rear-delt EMG differences between them.
So rows give your rear delts real work, but less than a rear-delt raise in the one model that compared them. Keep your rows and add direct sets. For mid-back rowing, see upper back exercises; for lat-focused pulls, see lat exercises.
Arm path, grip and load for rear delt flyes
Move your arms in a wide arc out to the sides, close to the 100°, thumbs-up position that gave the highest rear-delt signal in the fine-wire study. That study did not compare arm heights, so use the height where you feel the back of your shoulder work. Keep a soft, fixed bend in the elbows. If the elbows bend and straighten, the fly turns into a light row and the lats and biceps take over.
On a reverse fly machine, an EMG study in 19 trained men measured slightly higher mean rear-delt EMG with a neutral grip (palms facing each other) than an overhand grip, with a 95% CI of 0.1–7.4% MVIC. That difference is small and may be close to zero, so use neutral handles only if they feel comfortable.
Keep your shoulder blades fairly still and let your arms do the moving. That is a coaching cue, not a tested rule: in the seated-row EMG study in 14 men, holding the shoulder blades still gave modestly higher concentric rear-delt signal, but it also raised mid-trap and lat signal during the lowering phase, and no study has tested the cue in a fly.
Light weights are fine. A 2017 meta-analysis of 21 trials reported similar muscle growth with light loads (≤60% 1RM) and heavier loads (>60% 1RM) when sets went to failure, while heavy loads built more 1RM strength. A 2021 network meta-analysis of 28 studies in 747 adults likewise saw no clear growth difference between low, moderate and high loads with sets to failure. Neither covers light sets stopped far from failure, so a light fly pays off only when the last reps are hard.
Sets, reps and weekly volume
| Exercise | Sets per session | Reps | Effort |
|---|---|---|---|
| Reverse pec deck | 2–4 | 12–20 | 0–2 reps in reserve |
| Cable rear-delt fly | 2–3 | 12–20 | 0–2 reps in reserve |
| Head-on-bench dumbbell raise | 2–3 | 12–25 | 0–2 reps in reserve |
| Face pull | 2–3 | 12–20 | 1–2 reps in reserve |
| Wide-elbow rear-delt row | 2–4 | 8–15 | 1–3 reps in reserve |
| Band pull-apart | 2–3 | 15–30 | Close to failure |
A 2026 meta-regression of 67 studies estimated that muscle growth keeps rising with more weekly sets, with diminishing returns. It also found little evidence that spreading the same sets over more days adds growth. It gives no single best number, so start with 6 to 8 direct rear-delt sets a week over two sessions, and add 2 sets when progress stalls and you recover well.
That meta-regression counted sets that train a muscle indirectly as half a set, because that method fit the data best. As a rule of thumb, 8 hard row sets a week add about 4 rear-delt sets. That is bookkeeping, not a value measured for this muscle. Trainer tracks weekly sets per muscle, and sets per muscle per week covers the topic in depth.
On effort, a 2022 meta-analysis of 15 studies concluded that training to failure was not required for growth or strength, and did not harm them either. A 2024 meta-regression estimated that growth tends to rise as sets end closer to failure, though the authors call the exact relationship unclear. A missed rep on a fly is low risk, so end most sets 0 to 2 reps short of failure, rated with the RPE chart.
Technique cues for the top picks
Reverse pec deck
- Set the seat so the handles sit at shoulder height or slightly below.
- Take the neutral handles with a light grip and soft elbows.
- Sweep your arms back in a wide arc until they line up with your torso.
- Pause for one second, then take two to three seconds to return.
Cable rear-delt fly
- Set both pulleys at about shoulder height and cross the cables: left pulley in your right hand.
- Step back until the stacks lift at the start of the rep.
- Pull your hands out and back, not down, keeping the elbow bend fixed.
- Stop when your arms reach your sides, and keep your torso still.
Head-on-bench dumbbell rear-delt raise
- Rest your forehead on the top of an incline bench and hinge until your torso is close to parallel with the floor.
- Let the dumbbells hang under your shoulders, palms facing each other.
- Raise them out to the sides until your upper arms reach torso level, thumbs turned slightly up.
- Lower under control without bouncing at the bottom.
Face pull
- Set a rope at upper-chest to face height.
- Pull toward your face and split the rope ends apart.
- Finish with your elbows high and hands beside your ears.
- Keep your ribs down rather than leaning back to move more weight.
Regressions and progressions
| Exercise | Easier version | Harder version |
|---|---|---|
| Reverse pec deck | One arm at a time, lighter stack | One-second pause, three-second return |
| Cable rear-delt fly | Single arm, free hand braced | Pause with arms at your sides |
| Head-on-bench raise | Lighter dumbbells, shorter range | Standing rear delt fly without head support |
| Face pull | Band face pull | Heavier stack, two-second hold |
| Wide-elbow row | Cable rope rear-delt row | Heavier barbell rear-delt row |
| Band pull-apart | Lighter band, wider grip | Heavier band, narrower grip |
Add reps before load. Small dumbbells jump in big percentages: 5 kg to 7.5 kg is a 50% increase. Work up through 12 to 20 reps, add pauses or slower lowering, then move up. The progressive overload calculator can plan those steps, and resistance band back exercises has more band options for home.
A sample rear delt workout
Add this block to the end of a pull day. It takes about 15 minutes, with one heavier row before the isolation work.
| Exercise | Sets | Reps | Rest | Note |
|---|---|---|---|---|
| Barbell rear-delt row | 3 | 8–12 | 2 min | Wide grip, elbows flared, bar to upper chest |
| Reverse pec deck | 3 | 12–20 | 90 s | Neutral grip, last set near failure |
| Face pull | 2 | 12–20 | 60–90 s | Hands finish beside your ears |
| Band pull-apart | 1 | 20–30 | None | Optional finisher |
On a second day, such as a push day, add 3 sets of the cable rear-delt fly or the head-on-bench raise, for 8 to 9 direct sets plus the row credit. A full pull day workout shows where the block fits, and trap exercises covers the upper traps.
Common mistakes
- Swinging your torso to lift heavier dumbbells, which skips the part of the rep the rear delts should do.
- Bending and straightening your elbows during flyes, which turns them into light rows.
- Shrugging toward your ears, which hands the rep to the upper traps.
- Assuming presses cover the rear delts; the EMG data above show they barely use them.
- Stopping light sets far from failure, which wastes the main advantage of high-rep work.
- Pushing through sharp shoulder pain. Stop the exercise and see a qualified clinician instead.
What the research cannot tell you yet
- No training study has measured posterior deltoid growth, so every ranking here rests on mechanism, EMG and modelling.
- Face-pull evidence is a single computer-modelling study in 10 men with estimated forces. No EMG or training study of the face pull exists in the evidence reviewed here.
- Grip and arm-path evidence for flyes is one small EMG difference in 19 trained men plus light rehabilitation exercises in 10 subjects.
- The pressing EMG study in 13 men tested no pulling or rear-delt exercise. It only shows that presses and chest flyes barely use the rear delts.
- The half-set credit for rows is a model choice across many muscles, not a value measured for the rear delts.
Frequently asked questions
What exercise grows rear delts the most?
No study has compared rear delt growth between exercises, so the data cannot name a single winner. On mechanism, a reverse fly is the strongest pick: the reverse pec deck, cable rear-delt fly or head-on-bench dumbbell raise. Each isolates horizontal abduction, keeps your torso still and lets you push sets close to failure. Pick the one you have and progress it.
Are rear delts part of the back or the shoulders?
The rear delts are part of the shoulder. They are the back third of the deltoid, next to the middle and front portions. You train them mostly by pulling and by moving the arm backward at shoulder height, so they usually work alongside your upper back. That is why many programs place them on back or pull days.
Why are my rear delts weak?
Usually because they get little direct work. Presses and chest flyes load the front of the shoulder, and rows with tucked elbows hand most of the pull to the lats. The rear delts are also small, so heavy, swinging flyes let other muscles do the job. Add six to eight controlled reverse-fly sets a week and track your reps.
Do rows work the rear delts?
Yes. Rows extend the shoulder, and the rear delts assist that motion. Flaring your elbows away from your torso moves more of the work onto them. Rows still share the load with the lats, mid traps and biceps, so most lifters do better with rows plus a few direct reverse-fly sets each week.
How often should you train rear delts?
Two or three times a week suits most people, because it splits six to ten sets into short blocks at the end of other sessions. When weekly sets are equal, the number of days seems to matter less than the total, as the volume section explains. Light flyes are easy to recover from, so frequent short blocks work well.
Sources
The studies and reviews this guide cites, and what each one supports. Activation (EMG) studies measure electrical signal, not muscle growth.
- Electromyographic analysis of the rotator cuff and deltoid musculature during common shoulder external rotation exercises (Reinold et al., 2004)
A prone reverse-fly position (arm at 100° of abduction, fully externally rotated) produced high posterior deltoid EMG (88% MVIC); 10 subjects, light rehab exercises, no growth outcome.
- Effect of hand position on EMG activity of the posterior shoulder musculature during a horizontal abduction exercise (Schoenfeld et al., 2013)
A neutral grip on a reverse fly machine produced slightly higher rear-delt EMG than an overhand grip; the difference is small and no growth was measured.
- Activation of the three deltoid muscle portions during common strengthening exercises: A systematic review (Rabello et al., 2024)
Pull-ups and inverted rows ranked among the highest posterior-deltoid activators; values were averaged across mixed EMG methods and say nothing about growth.
- Different Shoulder Exercises Affect the Activation of Deltoid Portions in Resistance-Trained Individuals (Campos et al., 2020)
Presses and chest flyes produced very low posterior deltoid EMG and the lateral raise more; no pulling or rear-delt exercise was tested, so it cannot rank rear-delt exercises.
- Biomechanical evaluation of posterior deltoid exercises and kinematic trajectory modeling (Zhu et al., 2025)
The only face-pull study located: modelled rear-delt force was similar for face pulls and bent-over raises and higher than for a dumbbell row; forces were estimated, not measured.
- High-Density Surface Electromyography Excitation of Prime Movers Across Scapular Positions in the Seated Row (Padovan et al., 2025)
Fixed vs free shoulder blades in a seated row gave broadly similar back EMG, with modestly higher concentric rear-delt signal when fixed; one session, EMG only.
- Impact of different ranges of motion in the prone barbell row on muscle excitation (Fischer et al., 2025)
Changing range of motion in the prone barbell row did not change rear-delt EMG; one session, EMG only, no growth outcome.
- Longing for a Longitudinal Proxy: Acutely Measured Surface EMG Amplitude is not a Validated Predictor of Muscle Hypertrophy (Vigotsky et al., 2022)
Acute surface EMG amplitude is not a validated predictor of growth, so the EMG comparisons in this guide are not growth rankings; opinion review with no new data.
- Strength and Hypertrophy Adaptations Between Low- vs. High-Load Resistance Training: A Systematic Review and Meta-analysis (Schoenfeld et al., 2017)
With sets to failure, light and heavier loads built similar muscle size while heavy loads built more 1RM strength; it does not cover light sets stopped far from failure.
- Resistance Training Load Effects on Muscle Hypertrophy and Strength Gain: Systematic Review and Network Meta-analysis (Lopez et al., 2021)
No clear growth difference between low, moderate and high loads with sets to failure; mostly short trials, and no detected difference is not proof of equivalence.
- The Resistance Training Dose Response: Meta-Regressions Exploring the Effects of Weekly Volume and Frequency on Muscle Hypertrophy and Strength Gains (Pelland et al., 2026)
Growth rises with weekly sets with diminishing returns, frequency adds little at equal volume, and indirect sets fit best at half credit; it gives no optimal set number.
- Effects of resistance training performed to repetition failure or non-failure on muscular strength and hypertrophy: A systematic review and meta-analysis (Grgic et al., 2022)
Training to failure was not required for growth or strength and did not harm them; only 7 hypertrophy studies, all in young adults.
- Exploring the Dose-Response Relationship Between Estimated Resistance Training Proximity to Failure, Strength Gain, and Muscle Hypertrophy: A Series of Meta-Regressions (Robinson et al., 2024)
Growth tends to rise as sets end closer to failure; exploratory, with reps in reserve estimated by the authors and no exact optimal value.
Try the numbers for yourself.
Your next working set
Decide when to add reps, increase load, or repeat the weight.
RPE chart
Compare effort, reps in reserve, and estimated percentages.
A useful next read.
All guidesElbows-out rows and prone raises for the rhomboids and mid traps.
One vertical pull, one elbow-to-hip row, and the cues that help.
