The six-week race block gets sold as a precise recipe: weeks one to three, short 30-second efforts to lay down a mechanical reserve; weeks four and five, longer intervals with shrinking rest; week six, over-unders and race-pace pieces, then taper. It sounds like a law of physiology. It is mostly craft. The individual tools in that recipe are genuinely proven -- but a 63-rider trial that tested the exact thing the recipe insists on, the order of the intervals, found it made no difference at all. The block works. The precise choreography is one reasonable arrangement among many, not the one true path.
The Simple Version
A build block is real and it works, but you should know which parts are evidence and which are tradition. Proven: short intervals beat long ones for VO2max, 4x8-minute efforts are a gold standard, concentrating hard sessions into a block outperforms spreading them out, and over-unders keep you longer in the VO2max zone. Not proven: that you must run those tools in a specific order, that over-unders belong precisely in week six, or that six weeks is a magic number rather than a sensible convention. When researchers matched the total load and only varied the order, everyone improved the same amount. So build the block from the proven tools, aim for about six weeks because that's how long the adaptations take, run two quality sessions a week if you're an amateur, taper, and don't cook yourself in the last hard week.
How It Works
What's Law and What's Craft
The Tools Are Real
Start with what holds up, because a lot does. The components of a build block have solid randomized evidence behind them.
Short intervals beat long ones. Ronnestad's cyclists doing 30/15-second efforts raised VO2max 8.7% against 2.6% for those doing 4x5-minute intervals at matched effort. The mechanism is that the short rests never let oxygen uptake fall, so you bank more time in the red zone.
The 4x8-minute interval is a gold standard. Seiler compared it head to head: 4x8 min produced an 11.4% VO2peak gain, against roughly 5.5% for both 4x16 and 4x4. Accumulating 32 minutes at about 90% of max heart rate beat less time at higher intensity -- and felt easier.
Concentrating the stimulus works. Ronnestad's block-periodization group raised VO2max 4.6% while a traditional even-spread group changed nothing, and a 2025 study found a six-day short-interval microcycle lifted power at VO2max 3.7% versus 0.7% for regular training. Piling hard sessions together, then recovering, beats sprinkling them evenly.
And over-unders earn their place. Bossi showed that varying power within an interval -- surging up and settling back, the essence of an over-under -- kept riders above 90% VO2max for 410 seconds versus 286 for a steady effort, at the same perceived effort. As a way to accumulate time at high oxygen uptake and to rehearse the surges of a real race, they're legitimate.
So the toolbox is not folklore. Every item in it has a trial behind it.
The Order Is Not
Here is where the recipe overreaches. It insists you must progress in a specific sequence -- short efforts first to build a reserve, then lengthen them, then over-unders at the peak. That specific claim has been tested directly, and it failed.
Sylta and colleagues took 63 well-trained cyclists and ran three groups through twelve weeks of the same interval sessions in different orders: shorter-to-longer, longer-to-shorter, and mixed. Every group improved VO2peak and 40-minute power by 5 to 10%. Between the groups, nothing -- no significant difference, similar individual response patterns across all three. When the total load is the same, the order you arrange it in does not decide the outcome.
That doesn't make the recipe wrong to follow. Starting easier and building toward race specificity is sensible, motivating, and manages fatigue well. But it's a reasonable organizing principle, not a physiological requirement, and you should not believe you've ruined a block by doing the over-unders in week four instead of week six.
Why Six Weeks
The number six isn't magic either, but it isn't arbitrary. It tracks how long adaptations actually take. In a time-course study, exercise tolerance and peak power rose within two weeks -- fast neuromuscular and recruitment gains. Anaerobic threshold moved by four weeks, as the peripheral machinery of capillaries and enzymes caught up. And VO2max only shifted by six weeks, because it waits on the slower structural remodeling of the heart. (That study was in older adults, so read the exact weeks as a guide, not a stopwatch.)
Stack those and six weeks is the length that lets the whole cascade play out -- a couple of weeks to sharpen the neuromuscular side, a few more to move the threshold, and enough total time for VO2max to actually climb -- before you taper. It's the interval that fits the biology, which is exactly why it became the convention. It also, incidentally, explains the recipe's instinct to do neuromuscular-flavored short efforts early and metabolic work later: that ordering mirrors the natural sequence of adaptation, even though Sylta showed you don't have to enforce it.
Where Over-Unders Actually Sit
One honest caveat on the block's supposed centerpiece. Over-unders demonstrably raise your time above 90% VO2max (Bossi), which is a real training benefit. But there is no trial showing they improve race results more than ordinary threshold work, or that placing them in week six specifically is optimal. Their value is specificity -- they rehearse the stochastic surges of a real race -- not a unique molecular effect. Use them because they simulate racing and pack in high-intensity time, not because a law says the block must peak with them.
Example
Example: A Six-Week Block for an Amateur
Take a rider with an FTP of 260 W, six to nine training hours a week, preparing for a 70.3. Here is a block built from the proven tools -- with the parts that are convention flagged as such.
| Weeks | Key sessions (2/week) | What's evidence, what's craft |
|---|---|---|
| 1-2 | 4x8 min ~100-105% FTP; one short-interval day (e.g. 3x 13x30/15 s) | 4x8 and short intervals are proven tools (Seiler, Ronnestad) |
| 3-4 | Continue 4x8 / short intervals, nudge volume of work up | Progression is sensible; the exact order is not required (Sylta) |
| 5 | Over-unders (e.g. 3x 9 min, 2 min at 105% / 1 min at 90% FTP) | Over-unders raise time >90% VO2max (Bossi); week placement is craft |
| 6 | Taper: cut volume ~40-60%, keep intensity and frequency | Taper is well established (see the taper article) |
Two things the amateur must get right, and neither is the choreography.
First, session frequency. The elite block studies concentrate up to five hard sessions in a week. For an amateur juggling work, sleep debt and no massage therapist, that's a fast route to illness or injury. Two quality sessions a week plus one long ride is the durable dose. The block's benefit comes from adequate quality work absorbed consistently, not from copying an elite shock microcycle.
Second, don't overcook week five. The most useful warning sign is counterintuitive: in overreaching triathletes, heart rate recovery got faster -- by about 8 bpm -- even as performance dropped 2.1%. A heart that suddenly rebounds quicker while your numbers sag is not a sign of fitness arriving; it's a sign you've dug too deep. If that shows up, cut the last hard week short and slide into the taper early. The taper is not the reward for surviving the block; it's part of the plan.
The target power for the race itself -- roughly 80-87% of FTP for a 70.3, lower if you'll be out there over three hours -- is coaching convention drawn from race-file analysis, not a number a trial has pinned down. Treat it as a starting estimate to confirm in your long sessions, not a law.
Practical Rules
Practical Rules
-
Build from the proven tools, not the exact recipe. Short intervals, 4x8-minute efforts, concentrated blocks and over-unders all have real evidence. The specific order you run them in does not decide the outcome, so choose an order that's sustainable and motivating rather than agonizing over the sequence.
-
Aim for about six weeks because that's how long it takes. Neuromuscular gains come in two weeks, threshold in four, VO2max in six. Six weeks lets the whole cascade run before you taper. Shorter and you clip the VO2max gain; much longer and you're just accumulating fatigue.
-
Amateurs: two hard sessions a week, not five. The elite five-session shock block is not built for someone with a job and a sleep deficit. Two quality interval days plus a long ride, absorbed consistently for six weeks, is the dose that works without breaking you.
-
Progress by sensible defaults, and don't panic about precision. Starting with 4x8 and short intervals, then adding over-unders later, is a fine arrangement -- but it's a default, not a mandate. Doing the over-unders a week earlier or later changes nothing the evidence can detect.
-
Use over-unders for what they actually do. They pack in time above 90% VO2max and rehearse race surges. That's the reason to include them -- not a belief that a block must climax with them. They don't have evidence of beating plain threshold work for race outcome.
-
Watch for the paradoxical warning sign in week five. If your heart rate recovers faster than usual while your power at a familiar effort sags and fatigue climbs, that's functional overreaching, not form. Cut the last hard week and start the taper early rather than pushing through.
-
Taper -- it's part of the block, not optional. The last week or two should drop volume substantially while holding intensity and frequency. The gains you built in the block only show up if you arrive fresh. Details are in the taper article; don't skip it to squeeze in one more heroic session.
Evidence Base
Evidence Base
The components are on firm ground. Ronnestad (2015) on short-versus-long intervals, Seiler (2013) on 4x8, Ronnestad (2014) on block periodization, Solli (2025) on the shock microcycle, and Bossi (2020) on varied-intensity intervals are all randomized or crossover designs with clear effects. The samples are small -- 14, 16, 19, 35 riders -- so the exact percentages are indicative, but the direction of each is consistent and reproduced across the Norwegian group's body of work.
The crux finding is the strongest single piece. Sylta (2016) is 63 cyclists, the largest of the set, and it tested precisely the claim the popular recipe rests on: that interval order matters. It doesn't, at matched load (P > 0.05). This is the evidence that turns "the six-week block is a fixed sequence" into "the six-week block is a toolbox you can arrange sensibly." It is one study, in trained cyclists, over twelve weeks -- but it directly addresses the question, which is more than the recipe's precision can claim.
The six-week rationale leans on adaptation kinetics, with a caveat. Herrod (2020) gives the clean tolerance-then-threshold-then-VO2max sequence, but it was conducted in older adults, so the specific 2/4/6-week markers should be read as a pattern rather than exact timing for a trained athlete. The qualitative order -- neuromuscular before metabolic before cardiac -- is well supported elsewhere and is what justifies the roughly six-week length.
The overreaching marker is useful and slightly counterintuitive. Aubry (2015) is a prospective study of 31 triathletes and its paradoxical finding -- faster heart rate recovery accompanying worse performance -- is worth internalizing precisely because it's the opposite of what most athletes assume a quick-recovering heart means. Treat it as a flag to check against performance, not a standalone diagnosis.
What's genuinely unproven, and stated as such. There is no trial showing that a specific week-by-week choreography beats another arrangement, that over-unders in week six improve race results more than threshold work, or that the exact target-power percentages for each distance are correct to the point. Those are coaching conventions -- reasonable, widely used, drawn from experience and race data, but not established by controlled comparison. The article treats them that way. Two limitations run throughout: the interval research is almost entirely in trained men on bikes, so translation to running and to women is by inference, and the whole field studies six-week-ish blocks in isolation rather than across the multi-season arc where their real value would show.