Four minutes is genuinely enough, but only at an intensity almost nobody actually reaches.
A Tabata is eight rounds of twenty seconds of work separated by ten seconds of rest. Four minutes end to end, including the rests. Unlike EMOM or AMRAP it isn't an acronym: it's a surname, from the Japanese researcher whose work the interval structure is named after.
That naming matters more than it sounds, because the name travelled and the protocol didn't. What made the original interesting was not the clock. It was the intensity: work efforts far above the pace you could sustain aerobically, performed on a cycle ergometer where that intensity can actually be produced and measured. Twenty on and ten off with any movement at any effort borrows the timing and leaves the part that did the work behind.
The ten-second rest is the entire design. It's long enough to stop moving and short enough that nothing meaningful recovers, so each round starts more depleted than the last. The oxygen debt accumulates rather than clearing, which is why the fourth minute of a real Tabata is a different experience from the first.
The measure of a Tabata is consistency, not total output. Count the reps you complete in round one and treat that as your target for all eight. If round one gives you twelve and round eight gives you five, you didn't do a hard Tabata. You did one hard round and seven rounds of fading, which is a different and much less useful stimulus. Our own seeded Tabata carries the instruction directly: count reps in round one, then chase that number each round.
Movement choice is the whole game. It has to be something you can drive hard with no technical thought and no setup, from a standing start, with a rep that's identical at second one and second nineteen. Air bike, echo bike, rower, burpees, and not much else. The moment a movement requires you to think, the twenty seconds is spent thinking.
Four minutes is short enough that people naturally stack them. Three Tabata blocks with different movements is a twelve-minute session, and that's a common and reasonable way to run it, but it's three protocols rather than one.
Take our seeded air bike Tabata: eight rounds of twenty seconds sprinting, ten seconds off, four minutes total, scored in calories. Round one at full effort might give you nine calories. The honest target for the remaining seven rounds is nine. What actually happens on a first attempt is 9, 8, 7, 6, 6, 5, 5, 4, which totals fifty and feels dreadful. The disciplined version is 7, 7, 7, 7, 7, 7, 6, 6, which totals fifty-four, feels worse in a different way, and is a substantially better piece of training. Your score is the lowest round, not the sum. That's the whole read. Anyone can produce one good twenty-second effort. The protocol is asking whether you can produce eight, and the honest answer on a first attempt is usually no.
Tabata: 8 × (20 sec Assault Bike sprint) + 10 sec rest
The twenty and the ten are fixed in the timer rather than read from the prescription, because a Tabata that isn't twenty and ten isn't a Tabata. The player counts the work down, flips into rest by itself, and flips back. There's nothing to tap, and the rest cannot be skipped, which is deliberate: skipping the ten seconds is skipping the protocol.
The current round is labelled as an interval rather than a round, and each completed twenty-second effort is logged as its own split so the record shows eight entries rather than one four-minute block. On the final rest the player says so explicitly, so the last ten seconds don't feel like a mistake.
A multi-station Tabata authored as several four-minute blocks is expanded into its real interval count from the prescribed duration, and the movements rotate one per interval. So a twelve-minute three-station Tabata runs as twenty-four intervals with the right movement showing on each, rather than as three untimed lumps you have to track yourself.
The papers behind the empirical claims on this page, with the population each one studied. Where we could confirm only what a study set out to measure and not the size of what it found, the note says so rather than borrowing the paper's authority for a number nobody read.
I Tabata et al. (1996). Effects of moderate-intensity endurance and high-intensity intermittent training on anaerobic capacity and VO2max.
Two training experiments on a mechanically braked cycle ergometer. The moderate-intensity arm trained at 70% VO2max, 60 min/day, 5 days/week for 6 weeks. Anaerobic capacity was measured as maximal accumulated oxygen deficit.
Adult male subjects, mechanically braked cycle ergometer.Eight rounds of 20 seconds of work and 10 seconds of rest, four minutes in total. It is named after the researcher associated with the original interval study rather than being an acronym. The defining feature is not the clock but the intensity: the work efforts are meant to be far above any pace you could sustain, which is why it works best on a bike or rower where that effort can actually be produced.
Four minutes is enough to be a real stimulus, and it is not enough to replace a training week. It also is not free: a genuine Tabata generates fatigue out of proportion to its length, so it should be budgeted like a hard interval session rather than treated as a bonus. If the four minutes are not genuinely maximal, it is four minutes of moderate work and should be described as such.
Anything you can drive at full effort with no setup and no technical thought, and that looks the same in second 19 as in second 1. Air bike, echo bike, rowing machine, ski erg and burpees are the reliable choices. Loaded or skill-dependent movements, especially anything overhead, are a poor fit because the structure removes exactly the attention those movements need.
By your worst round, not your total. Count reps or calories in round one and try to repeat that number seven more times. A descending sequence means round one was faster than your true repeatable output. Holding a slightly lower number for all eight rounds is both a better score and a better piece of training.
Build a plan and Fire Your Coach will schedule the structure, run the clock, and record what actually happened round by round, so next time the prescription starts from your real numbers.