The sit-up has been the default core exercise for close to a century. It is also the exercise spine biomechanists are most cautious about. The reasons are real, but they are more specific and more conditional than most fitness content admits, so they are worth walking through properly.
What the lab actually measured
Stuart McGill's group at the University of Waterloo spent decades measuring loads on the lumbar spine during exercise. Their models estimated compressive forces of roughly 3,000 to 3,400 newtons during a traditional sit-up, varying with technique, speed and the biomechanical model used.
It is worth being precise about what that number represents. It is an internal force, generated mostly by muscles contracting to flex the trunk, transmitted through the spine. Every hard muscular effort produces internal compression. This is a biomechanical magnitude, not a load you are carrying on your back.
The NIOSH comparison, and its limits
The US National Institute for Occupational Safety and Health uses 3,400 N of compression at L5/S1 as one of the design criteria in its revised lifting equation. It is the threshold above which repetitive occupational lifting is associated with higher rates of low-back injury.
The comparison is informative, and it is routinely over-read. NIOSH is describing workers performing manual lifting for hours a day across a working life. It is not a per-repetition danger threshold, and it was never designed to evaluate three sets of fifteen sit-ups. The two situations share a number. They do not share a context.
What the flexion research shows, and what it does not
McGill's group also demonstrated that repeated flexion under compression can produce annular fissures, disc protrusion and herniation. That work is important, and it is the basis for most of the caution around high-volume spinal flexion.
It also carries a limitation worth stating plainly: much of the research on this mechanism was performed on porcine spine specimens and cadaveric tissue, not on living people in training. Isolated tissue cannot adapt, remodel, rehydrate between sessions, or be protected by active musculature. In a living human, the relationship between repeated flexion and injury is mediated by recovery, training adaptation, muscle quality, disc hydration and genetics.
So the accurate claim is that repeated loaded flexion is a plausible risk factor worth managing. The inaccurate claim — the one we are not making — is that sit-ups cause disc herniation.
What sit-ups train, and what they train less
A sit-up is spinal flexion driven primarily by the rectus abdominis and, past roughly thirty degrees, increasingly by the hip flexors: the iliopsoas and rectus femoris. That handoff is well established.
On the deeper stabilisers, precision matters. EMG shows the obliques and transversus abdominis do activate during a crunch — the obliques moderately, the transversus in a stabilising role. They are not idle. They are simply recruited less than in movements built specifically to challenge them, such as planks and anti-rotation work.
There is also an unhelpful irony in anchoring the feet. It lets you perform more repetitions, but it does so by increasing hip flexor involvement. The psoas attaches directly to the lumbar vertebrae, so more psoas means more compressive load on precisely the segment in question. That makes the feet-anchored version harder on the lower back, not safer.
What modern core training emphasises
Over the last fifteen years the field has shifted toward training the trunk to resist movement as well as produce it, since resisting is mostly what the trunk does while walking, running, carrying and lifting. Three categories:
- Anti-extension — resisting the lower back arching. Planks, dead bugs, ab wheel rollouts.
- Anti-rotation — resisting a twisting force. Pallof presses and single-arm carries.
- Anti-lateral flexion — resisting side bending. Side planks, suitcase carries.
McGill's recommended trio is the modified curl-up, the side plank and the bird dog. One clarification that gets missed constantly: his curl-up is not a traditional sit-up. The hands go under the lower back to preserve its natural curve, and the movement is small — the head and shoulders lift only slightly. The point is to load the abdominal wall while keeping the spine out of end-range flexion. Substituting a traditional sit-up defeats the purpose of the exercise.
It depends what you are training for
This is where blanket answers fall apart, and where most content on this topic goes wrong.
If your goal is spinal stability and moving without pain, anti-movement work is the better-supported choice — and that is what most people mean when they say they want a stronger core.
If your goal is rectus abdominis hypertrophy, flexion exercises are among the highest activators of that muscle, which is why they remain a staple in bodybuilding programmes. Planks are not superior for that goal. They serve a different purpose.
Sit-ups are not the least evidence-backed exercise in the gym. They are highly effective at activating the rectus abdominis. They are simply not the best tool for trunk stability, and their loading pattern rewards moderation.
So should you never do sit-ups?
No — and the honest position is less dramatic than either extreme.
A healthy spine tolerates flexion; bending forward is a normal human movement. The evidence argues against very high-volume repeated flexion as your only core training, not against the movement existing.
- If you currently have low-back pain, a history of disc problems, or degenerative changes, avoid loaded spinal flexion and check with a physiotherapist first.
- If your back is healthy, a moderate number of well-executed sit-ups is unlikely to harm you. Sets of fifty a day for years with sloppy technique is a different story.
- Whichever you choose, anti-extension and anti-rotation work deserve a place in the programme rather than being an afterthought.
Training this at home
Resistance tubing is particularly good for anti-rotation work, because the resistance pulls horizontally and your job is to refuse to turn.
The tube system on a multi-function ab trainer is genuinely useful here: anchored to one side and pressed straight out from the chest, it becomes a Pallof press. The same unit gives you seated rows, curls and leg presses. Those are the uses we would build a routine around. The assisted sit-up function works, and if your back is healthy there is nothing wrong with using it — keep the volume moderate and the tempo controlled.
A set of fabric resistance bands covers the rest: banded dead bugs, side-lying hip abduction for the lateral chain, and glute bridges, which train hip extension without asking the lumbar spine to do work the glutes should be doing.
A balanced core session
Three times a week, twelve minutes:
- Modified curl-up — 3 × 8, hands under the lower back, small range
- Side plank — 3 × 20–30 s per side
- Bird dog — 3 × 8 per side, slow
- Pallof press with tubing — 3 × 10 per side
- Suitcase carry — 3 × 30 s per side, one heavy object, walk tall and refuse to lean
Carries are among the most practical anti-lateral-flexion exercises there are, and they need no equipment beyond something heavy with a handle.
Progress by adding time to the holds and resistance to the presses before adding flexion repetitions.
The short version
Sit-ups are not dangerous. They are a flexion exercise carrying real load, with strong evidence for one specific outcome and weaker evidence for the outcome most people actually want. Anchoring your feet makes the exercise harder on your back, not easier.
Build your core work around positions that keep the spine neutral, include flexion in the programme if your back is healthy and you want it, and stop treating any single exercise as either essential or forbidden.
See the Home Gym Starter collection →
Keep reading
- Can You Build Real Upper-Body Strength Without Weights? — what a meta-analysis comparing elastic resistance with dumbbells and machines actually found.
- How Long Should You Actually Stretch? — the numbers on duration and weekly volume, including one that surprises most people.
- Pilates Bar vs Reformer: What Actually Works in a Small Apartment — an honest comparison for training in tight spaces.
Sources
- McGill S. Low Back Disorders: Evidence-Based Prevention and Rehabilitation. Spine Biomechanics Laboratory, University of Waterloo. uwaterloo.ca
- Waters TR, Putz-Anderson V, Garg A. Revised NIOSH Equation for the Design and Evaluation of Manual Lifting Tasks. The 3,400 N L5/S1 compression value is one of three design criteria. cdc.gov/niosh
- Schoenfeld B. Core Training: Evidence Translating to Better Performance and Injury Prevention. Strength and Conditioning Journal, 2010.
- The McGill Big 3 for core stability. squatuniversity.com
This article is general information about exercise, not medical advice. If you have current or recurring low-back pain, consult a physician or physiotherapist before starting or changing a core training programme.