Getting abs is two separate jobs, and most people run one of them while measuring the other. The first job builds the muscle, which happens the way every other muscle grows: loaded sets, taken close enough to hard, repeated for months. The second lowers body fat far enough for that muscle to show, which happens through a calorie deficit. Ab training does the first job well and the second job not at all, so a year of crunches with no deficit changes what is under the skin and nothing about what you can see.
That split is also why the two jobs need different instruments. Training progress reads as a heavier or longer set of a named movement over time. Fat loss reads as a waist measurement next to a bodyweight trend. ChatRPE is an AI coach built on a Universal Logbook, so both of those live in one history and can be drawn on one chart, which turns "is any of this working" into a question with a picture attached instead of a guess in front of a mirror.
The short version
- Core muscles respond to load and progression like every other muscle, and almost nobody gives them either.
- Visibility is a body fat outcome. No exercise selection changes where fat leaves, which the belly fat guide covers in full.
- Holds are measured in seconds and loaded movements in reps and weight. Logging them in the same units is what makes core training unreadable later.
- Two to four hard sets, twice a week, across two or three core patterns is a normal starting dose.
- Track the training with a strength line and the fat with a tape measure. Judging either one by the mirror wastes months.
Why ab training almost never progresses
Ask someone what they bench and you get a number that has changed over the years. Ask what they do for abs and you get "three sets of crunches," which is the same answer they would have given in 2019. That is the whole problem. Nobody would expect a bench press to grow from the same unloaded set of 20 repeated indefinitely, yet that is exactly how core work is usually run.
Two habits keep it stuck. Core work lands at the end of the session, when the willingness to add weight has run out, and most of the popular movements have no obvious place to put weight, so the only lever anyone reaches for is more reps. Past about 20 reps a set stops being strength training and starts being a conditioning drill you happen to feel in your stomach.
The fix is a decision rather than a better exercise. For each movement you keep, name the number that is supposed to go up, then make it go up.
The four core patterns and what each one is measured in
Your midsection resists movement more than it creates it, so useful core training is organized by what it resists. Four patterns cover the job, and each one has its own natural unit.
| Pattern | Example movements | How it gets harder | What to log |
|---|---|---|---|
| Anti-extension | Ab wheel rollout, hollow hold, plank variations | Longer lever, added load, longer hold | Reps, or seconds held |
| Flexion under load | Cable crunch, weighted decline sit-up, hanging leg raise | More weight on the stack or in the hands | Weight and reps |
| Anti-rotation | Pallof press, half-kneeling cable chop | Heavier cable, longer distance from the anchor | Weight and reps |
| Anti-lateral flexion | Suitcase carry, side plank, single-arm farmer's walk | Heavier carry, longer distance or hold | Load with seconds or distance |
Pick one movement from two or three of those rows and run them the way you run any other lift. A cable crunch at 40 pounds for 12 becomes 45 for 12 in a few weeks. A hollow hold at 30 seconds becomes 45. An ab wheel rollout from the knees becomes one from a longer starting position. The progressive overload guide covers reading that trend across sessions rather than reacting to one good day.
The unit matters more than it sounds. A plank is a time-domain exercise and a cable crunch is a load-domain one, and writing both into your history as "abs, 3 sets" throws away the only number that could have told you whether either improved.
How many sets of core work per week
Treat it as a muscle group with the same arithmetic as any other. Two to four hard sets, twice a week, is a sensible starting dose for most people, and it goes up only if that stops producing anything. The weekly sets guide covers how to count what you are actually doing, which for core work is usually less than people think, because half the sets are warm-up-grade effort taken at the end of a session.
Heavy squats, deadlifts, carries, and overhead pressing all load the trunk hard, so a lifter running those is not starting from zero. Direct core work adds the ranges those lifts never reach, mainly loaded flexion and long-lever anti-extension. If your program already has carries and heavy hinging in it, two direct sets twice a week is plenty of addition.
Log core work like the lift it is
Say the session in one sentence and it lands as structured data with the numbers separated: "cable crunch 3 sets of 12 at 40 pounds, then three hollow holds at 45, 40 and 38 seconds." ChatRPE splits that into named exercises with per-set values, and because entry shapes follow what you actually say, a set measured in seconds is stored as seconds rather than squeezed into a reps column. Say it out loud between sets if your hands are chalky; the hands-free logging guide covers dictating a set without unlocking your phone.
Keep the movement name stable, because filtering is literal and "hanging leg raise" and "leg raises" become two lines. The home workout guide goes deeper on naming variations when the load lives in the exercise rather than on a bar.
Once the names are stable, three requests carry the whole training readout:
- Ask for the heaviest set of one named core movement over time and you get a line drawn from the weight and reps already in your history, with nothing calculated by hand or saved as an extra metric.
- Ask for the longest hold you logged for a named movement and the seconds chart the same way the pounds do.
- Ask for a heatmap counting the sessions that included direct core work, which usually answers "why has nothing changed" before any of the other charts do.
Try this: "Chart my heaviest cable crunch set and my longest hollow hold over the last four months, and tell me whether either one has actually gone up."
What to measure while the fat comes off
The second job has one mechanism and one honest instrument. The mechanism is a calorie deficit you can hold for months, sized in the calorie deficit calculator, with protein high enough to keep the muscle you are trying to expose, which the protein calculator puts a band around. The instrument is a waist measurement taken at the same spot and time of day, read against your bodyweight trend rather than against yesterday.
There is no body fat percentage that puts abs on a given person. Where your body draws fat from, and in what order, is set by genetics, and the abdomen is commonly the last region to change even while everything else is going well. That is why the tape beats the mirror by a wide margin: half an inch off the waist is real and completely invisible day to day.
Log the tape in a sentence and it charts like any other metric. Then ask for both jobs on one picture:
Try this: "Put my waist measurement and my heaviest cable crunch set on the same chart with separate axes for the last six months, with a moving average on my bodyweight underneath."
Two axes let a number in inches and a number in pounds share one graph without either flattening the other. When the strength line climbs and the waist sits still, the training is working and the deficit is not, which is a completely different problem from the reverse. ChatRPE also watches for metrics that have stopped moving and raises them rather than waiting to be asked, so a flat quarter surfaces as a note instead of a realization in November. If both lines have genuinely been flat for weeks, the weight loss plateau guide works through the causes in order.
If a leanness target has ever been something you chased in a way that worried you or the people around you, set the target with a clinician or a registered dietitian and use the log to hold yourself to what they agree with you.
How long does it take to get abs
Two clocks, running at different speeds, and conflating them is why the question feels unanswerable.
The training clock is fast. Give a loaded core movement two months of honest progression and the number on it moves, the same way a press or a row moves. That change is fully measurable long before it is visible.
The visibility clock depends almost entirely on where you start. Someone within a few pounds of showing definition may get there in a couple of months of a modest deficit. Someone starting further out is looking at a longer stretch, and the arithmetic is the ordinary fat-loss arithmetic rather than anything specific to abs: a rate of roughly half a percent of bodyweight a week is a commonly used middle setting, and the total distance is what sets the timeline.
Treat those as planning estimates rather than facts about your body, which is exactly why the tape gets read every week or two. The results timeline guide covers which signals move first across every goal.
Frequently asked questions
Do ab workouts work?
For building the abdominal muscles, yes, on the same terms as any other training: loaded sets, progressed over months. For revealing them, no amount of core work substitutes for a calorie deficit, since fat leaves the whole body in an order your genetics set. Run both jobs and judge each one by its own number.
What are the best ab exercises?
The ones you can add load or leverage to, so the set gets measurably harder over time. A cable crunch, a hanging leg raise, an ab wheel rollout, a Pallof press, and a heavy suitcase carry cover the four patterns between them. Any movement stuck at unloaded sets of 25 has stopped being strength training whatever it is called.
How long does it take to get abs?
The strength side responds in weeks and the visible side depends on how much fat is covering the muscle, so the honest answer is months rather than a date, and the distance from your starting point sets it. Track a core lift and a waist measurement together and you will see one of them move well before the other.
Can you get abs without losing weight?
Sometimes, early in training or after a long layoff, when muscle is being added while fat comes off at the same bodyweight. That is the case where the scale sits still for weeks and the waist drops, and it is the reason a scale-only app reads a working plan as a failure. The body recomposition guide covers what that looks like in the numbers.
How many times a week should I train abs?
Twice a week, two to four hard sets each time, is a normal starting point, and consistency across months matters far more than the exact number. Core muscles recover quickly, so the limit is usually session time rather than fatigue.
Do I need to train abs directly if I squat and deadlift?
Heavy compound lifting genuinely trains the trunk, and plenty of strong people have never done a crunch. Direct work still adds loaded flexion and long-lever positions those lifts never reach, and it gives you a number to progress, which a brace inside a squat does not. Two sets twice a week is enough addition for most programs.
What is the best app for tracking core training?
The one that keeps a hold measured in seconds and a crunch measured in pounds as separate, chartable numbers, and can show either one next to your waist measurement. ChatRPE logs the session from a spoken or typed sentence, draws the heaviest set of a named movement straight from your history, and reads your training, food, and measurements together instead of leaving them in three apps.