2k Erg Training Plan: Eight Weeks to a Faster Piece

Eight weeks. Long enough to take real seconds off a 2k, short enough that no week in it can drift. This guide is the plan: what each week asks of you, the four session types and the job each one does, where the volume peaks and where it gets cut, and how to row the test at the end. It assumes you can read a split and know your zones; if either is new, the erg splits guide and the erg training zones guide cover them, and this page stays on the block itself.

Set the target before week one

Every session in this plan is keyed to one number, T, the 2k split you are training to hold, and you get it from a recent piece: enter your last 2k, or a hard 1k, 5k or 6k, into the 2k predictor to set your target split, then take one to two seconds per 500m off what it gives you. That is T. A 7:30 rower (1:52.5 per 500m) aiming at 7:22 has a T of 1:50.5.

Why so small? Because the erg prices ambition in cubes. Concept2's own formula puts watts at 2.80 divided by pace cubed, so two seconds of split at 1:50 cost about 15 watts while the same two seconds at 2:05 cost about 9. Roughly 1.5-2 percent is what trained rowers moved their 2k by in the studies below, over blocks of four to eleven weeks: for a 7:30 rower, 7-9 seconds, about two seconds of split. Newer rowers move more. Set T for the athlete you are today. Let the checkpoint in week four move it.

Why the block is mostly easy rowing

A 2k is aerobic work with a hard finish stapled to it. In race simulations on the ergometer, aerobic metabolism supplied about 84 percent of the energy, with the alactic and lactic systems splitting the rest, and that ratio decides the shape of the block: most of the minutes go to the aerobic engine. The sprint work is small, late and specific.

The best rowers in the world got to the same answer by trial. Between the 1970s and the 1990s, Norwegian international rowers raised their monthly low-intensity hours from 30 to 50 and cut high-intensity hours from 23 to about 7, while VO2max rose 12 percent and 6-minute erg performance nearly 10. And when national elite rowers spent 11 weeks with 93-94 percent of their training time in zone 1, polarized and pyramidal arrangements of the remaining hard hours each improved 2000m power by about 1.5 percent, and neither beat the other. The size of the easy share mattered more than how the hard share was cut.

It scales down. At four to five sessions a week the same principle comes out as two hard sessions, everything else steady, and the hard ones turning race-specific as the test approaches.

The four session types and the job each one does

• Steady state (UT2 and UT1). T + 18 to 25 per 500m, rate 18-22, 40-60 minutes. The aerobic base. Capillaries around the working muscle, and a few thousand low-rate strokes a week that groove the technique you will race with. Row it too hard and you take the edge off the two sessions that need it.
• Threshold (AT). T + 8 to 10 per 500m, rate 24-26, 20-32 minutes of work. Roughly your 6k pace. It raises the power you can hold at a blood lactate around 4 mmol, and that number is the middle thousand of the 2k. When 19 well-trained rowers swapped two of ten weekly sessions for 2.5-minute intervals at peak power for eight weeks, they gained more threshold power (effect size 1.15) and VO2max (0.95) than the all-aerobic group, with a small edge on the 2k time trial (0.25).
• VO2max intervals (TR). T + 2 to 4 per 500m, rate 28-30, 15-20 minutes of work. This is where the ceiling moves. The best-known erg protocol is Driller's: 8 x 2.5 minutes at 90 percent of the speed at VO2peak, seven sessions in four weeks, which produced greater improvements than conventional training in well-trained rowers: 1.9 percent on 2000m time, 5.8 percent on 2000m power, 7 percent on relative VO2peak, and weeks two, three and five use it almost as written. The VO2max intervals guide covers the format.
• Race pace (AN). T and faster, rate 30-34, 3-4 km of work in pieces. 500m and 750m repeats at the split and rate you will race, and they buy the anaerobic 16 percent and, more than that, the rhythm: what T feels like at rate 32 when the breathing is already short. Not before week six.

Strength stays in the block: two short sessions a week through week five (squat, a deadlift or Romanian deadlift, a row, a press), one in weeks six and seven, none in week eight. A meta-analysis of strength training in rowers found a small but significant effect on rowing-specific performance (SMD 0.32) and on lower-limb strength (0.42), and none on endurance capacity. A small lever. Pull it on hard days so the easy days stay easy.

The eight weeks at a glance

Read the table with T in your head. T + 10 means ten seconds per 500m slower than target, T minus 2 means two seconds faster, r24 means 24 strokes per minute, hard session A lands on Tuesday, hard session B on Friday, steady state fills Monday, Wednesday and Saturday, and Thursday and Sunday are off the erg. Weekly time counts warm-up and rest, rounded.

Steady-state lengths climb with the volume: 40-45 minutes in week one, 50-60 by week three, back to 40 in the deload, 50-60 through the build, 40-45 in week seven and 25-30 in test week. One session is allowed to change the plan, the checkpoint in week four, and its result goes into the predictor: if it says a faster 2k than the one you set T from, move T by the difference. If it says slower, hold T. Check your sleep before you decide anything.

How volume and intensity shift, and the ten-day taper

Weeks one to three build the base. Erg time climbs about 15 percent a week to a peak near four hours, the threshold reps lengthen from 10 minutes to 15, and the VO2 session settles into the Driller format. Nothing is at race pace yet.

Week four is the deload. Volume drops about 30 percent, both hard sessions stay but shrink, the checkpoint piece replaces Friday's session, and skipping the deload is the most common way to arrive at week six flat. The deload guide explains why.

Weeks five and six are the build. Volume returns to the peak and holds. The hard sessions turn toward the race while it does: longer and faster VO2 reps, threshold reps a second quicker, and the first race-pace session in week six at rate 30-32, with stroke rate climbing underneath, 18-22 steady, 24-26 threshold, 28-30 VO2, 30-34 race pace, so by test day T at rate 32 holds no surprises.

Week seven sharpens. Volume comes down about 20 percent. Intensity peaks. The 750m pieces are the hardest session of the block, and the last long steady state is Saturday.

The taper is the last ten days, from the Wednesday of week seven to the test. It is progressive: week seven's Friday and Saturday sessions are already shorter than the build's, and week eight halves everything. Two findings anchor it. A meta-analysis of 27 tapering studies found the largest performance gains from a two-week taper in which training volume fell 41-60 percent with no change to training intensity or frequency, and a 2023 meta-analysis of 14 studies in endurance athletes landed in the same place: 8-14 days worked best, 41-60 percent off the volume, intensity and frequency kept. The tapers that cut frequency did worse. Cutting intensity bought nothing. The tapers that followed an overload period gained the most, and week seven is that overload.

• Keep the sessions, shorten them. Week eight's total is about 1.75 hours against a four-hour peak. A cut of more than half. The Tuesday 4 x 500m keeps the intensity in the week.
• Touch T every session. Twenty strokes at race pace inside an easy row keeps the rate and the split in your hands without costing recovery.
• The last real piece is Tuesday. Three full days before the test. Thursday is a primer, not a workout. No lifting in test week; the last lift is Tuesday of week seven.

If your test date is fixed and the eight weeks do not line up, cut from the base weeks, never from the taper, because six weeks with a proper taper beats eight weeks that end tired. The tapering guide covers race tapers.

Test day protocol

• Same machine and drag factor, at the time of day you rowed the hard sessions. Set the monitor to a 2000m piece, show split and rate, and once the clock starts the pacing guide's four-phase plan for the 2k is the script.
• Eat three hours out. A normal carbohydrate meal. Nothing new. Around 500 ml of water two hours before.
• Warm up shorter and easier than instinct says. When 14 trained rowers compared a shorter, lower-intensity warm-up with their traditional one, the easier warm-up produced more mean power in a 10-minute all-out erg trial (322 versus 316 W), most of it in the first seven and a half minutes. Build yours like this: 10 minutes easy at rate 18-20, then 3 x 20 strokes building to race rate with a minute easy between, then 2 x 10 strokes at T, then 3 minutes easy. Sit two minutes. Go.
• First 500 at T, not faster. The most expensive mistake on the erg. Rate 30-32 through the middle thousand, lift from 1500m, empty the last 250, and every 20 strokes read the average split, not the current one.
• Afterward, ten minutes easy, then log it. Let the Concept2 logbook sync carry the piece into the app. That result is next block's T.
• If it goes wrong, do not retest tomorrow. A blown start is not fitness lost. Neither is a bad night's sleep. Repeat test week from Tuesday and go again in five to seven days.

Where eight-week blocks go wrong

• Steady state rowed at threshold. T + 12 feels productive. It is, at the cost of Tuesday and Friday. Hold T + 18 or slower, and the rate at 20.
• Race pace in week two. Pieces at T before the engine can support them teach a rhythm you cannot hold and a split you start to fear. Week six is early enough.
• Strength on easy days. A heavy squat session on Wednesday turns the easy day into a hard one and leaves nothing between Tuesday and Friday. Lift after the hard rows.

How Movement Rebels fits

The plan above is fixed on paper. No eight weeks are. Connect your Concept2 logbook and every erg session syncs into Movement Rebels alongside the rest of your training, and the coach sees the week where every session ran long, and the intensity distribution that says too little of it was easy. Give it the test date and the split you are chasing and it locks a dated goal, then reads where you are in the arc each week: which phase, which week of it, and whether this is a deload week.

The part a printed plan cannot do is the cross-check. Log Tuesday's squats in the calendar and they count toward the week's load, then let a Garmin or Apple Health sync show a resting heart rate six beats over your baseline on Friday morning, and the morning brief names it. One line in chat moves Friday's threshold session to Saturday.

Set T from a recent piece tonight, and put the test on the calendar eight Saturdays out. Then row Monday at T + 20.