Threshold and sweet spot intervals without hills: how to ride them on flat and rolling roads

Why flat and rolling roads suit threshold work

Threshold (95-105% of FTP) and sweet spot (88-94%) are long, steady, sub-maximal efforts. They need a stretch of road and a gear, not a gradient. On a 20-minute time trial, competitive cyclists averaged 284 W outdoors and 280 W indoors, a difference that was not significant [Jeffries et al. 2019].

The climb is a convenience, not a requirement. A steady grade holds your speed down and your resistance up, so power stays level with little thought. On the flat, air resistance does the same job: at threshold speed most of your power goes into pushing air, and that load is smooth as long as the wind is steady. The two settings feel different, but the dose in the legs is what the plan asks for either way.

The evidence for riding the same number outside is older than power-meter forums. Eight trained cyclists rode three 40 km time trials indoors and three on a local road course; mean power was 303 W indoors and 312 W outdoors, again not significantly different, and both settings were highly repeatable [Smith et al. 2001]. If you can hold 250 W for 20 minutes on a trainer, you can hold close to 250 W for 20 minutes on a flat road. If your outdoor number reads clearly higher or lower, that is a calibration and testing question; our guide to indoor vs outdoor FTP covers it.

This is why threshold is the sensible place to start doing intervals outside without hills. Short VO2 max reps and sprints suffer more from a crest, a junction or a gust, because the effort is too short to absorb a lost 20 seconds. A 12- or 20-minute rep at 90-100% of FTP absorbs small disruptions and still delivers its average.

Holding the range on rolling roads

Expect outdoor power to vary about twice as much as indoor power. In the same riders, the within-ride standard deviation was 69 W outdoors against 33 W indoors, at the same average [Jeffries et al. 2019]. So set a range, shift before the road changes, and grade the rep by its average and time in range.

On rollers, power is lost in two places. On the way up, riders surge: the grade bites, cadence drops, and they stand and push 30-40% over target for 20 seconds. Over the top and on the way down, they coast, and power falls to zero. Both cost the rep. The surge spends energy above threshold, and the coasting gives back time in range that the average then has to recover.

The fix is mechanical. Shift to an easier gear just before the road tilts up, so cadence stays near your normal 85-95 rpm and power rises only to the top of the range. Over the crest, do not stop pedalling: ease to the bottom of the range for a few seconds, then shift up one or two gears before the descent so the pedals still have resistance. On a descent too steep to hold power, accept the dip, and do not chase it back with a surge at the bottom.

A small amount of variation is not a failure. In a lab study, power that swung 15% above and below the average every two minutes produced higher oxygen uptake, heart rate and blood lactate than constant power at the same average [Kolsung et al. 2020]. The same authors found perceived exertion similar at the higher intensity and called variable-power training a viable alternative to constant power [Kolsung et al. 2020]. A rolling-road threshold rep costs a little more than the same average on a trainer. It does not count for less.

How to grade the rep, then: look at its average power first, then at how much of it sat inside the range. A 20-minute rep that averages 96% of FTP with 80% of its time in the 95-105% band did its job, even with a few spikes and dips. A rep with the right average but long stretches far above and below the range is the one to fix next time, usually by shifting earlier.

Headwind and tailwind legs

Wind is the hill of a flat road. Ride your threshold reps into the headwind or crosswind, where resistance is high and steady, and use the tailwind leg for recovery. When a rep must cross both, push slightly harder into the wind: about 5% above the average worked in a lab time trial [Atkinson & Brunskill 2000].

A tailwind is the hard part of flat-road threshold work. With the wind behind you, speed rises, air resistance falls, and on many bikes you run out of gears before you reach threshold power. A headwind does the opposite. It loads every pedal stroke like a long, gentle climb, and power is easy to hold steady. On an out-and-back road, check the wind before you leave, ride out into it with the reps, and spin home on the recoveries.

If the rep has to turn and run with the wind, do not try to hold one exact number across both directions. In a simulated 16.1 km time trial with a headwind first half and a tailwind second half, riders who rode 5% above their average into the headwind and eased off in the tailwind finished slightly faster than at constant power, with lower perceived exertion and lower lactate [Atkinson & Brunskill 2000]. A mathematical model points the same way: a small rise in power into headwinds and up grades, paid back on tailwinds and descents, saves time at the same average [Swain 1997].

The training lesson is narrow. These studies were about finishing time, not training response, and the variation was small. For an interval session, read them as permission to sit near the top of your range into the wind and near the bottom with it behind you. A 100-105% headwind and 95% tailwind split still averages threshold.

2x20, 3x12 or 4x8: pick the rep the road allows

Choose the rep length from the longest uninterrupted stretch you have. At about 35 km/h, an 8-minute rep needs roughly 4.7 km without a stop, a 12-minute rep about 7 km, and a 20-minute rep almost 12 km. Shorter reps lose little: in 35 cyclists, 4x8 minutes beat 4x16 minutes over seven weeks [Seiler et al. 2013].

Most riders without hills also lack a 12 km road with no lights, junctions or stop signs. That is where 4x8 earns its place. Seiler and colleagues trained recreational cyclists for seven weeks with 4x4, 4x8 or 4x16 minutes at the hardest effort they could sustain; the 4x8 group gained the most in VO2 peak and in power at 4 mmol lactate [Seiler et al. 2013]. That study tested intensities above sweet spot, so it does not prove 4x8 is the best threshold session. It does show that 32 minutes split into 8-minute reps is a strong dose, not a compromise.

A simple rule for your own roads: count the longest stretch you can ride without unclipping, and convert it to minutes at your threshold speed. Under 8 minutes, ride 5-6 x 6 minutes or move the session indoors. Between 8 and 12 minutes, ride 3x12 or 4x8. Over 15 minutes, a 2x20 fits, and a loop lets you string reps without a turnaround. Each turnaround costs 20-40 seconds of soft pedalling. Put it in a recovery, not in the middle of a rep. There is no study on stops and interval quality; this rule comes from the arithmetic of rep length, not from a trial.

Once the reps are long and the road varies, a head unit helps more than memory does. A structured workout on a Wahoo ELEMNT or Hammerhead Karoo shows the target range, the time left in the rep and the lap average, so you ride the range rather than chase one number. AdaptCycling sends each planned workout to both units with its power ranges, and it picks which session today is from the week you actually had, so a missed Tuesday does not leave Thursday's 2x20 unchanged. Its Suggested Routes feature, part of the paid plan, offers loops sized to today's ride time and checks each loop for straight stretches long enough for the reps; it picks a loop where the work fits, not the exact spot each rep begins.

Quick answers

Sources cited in this guide

1. 01
   Jeffries et al. 2019. An Analysis of Variability in Power Output During Indoor and Outdoor Cycling Time Trials. International Journal of Sports Physiology and Performance.
2. 02
   Smith et al. 2001. Reliability of mean power recorded during indoor and outdoor self-paced 40 km cycling time-trials. International Journal of Sports Medicine.
3. 03
   Kolsung et al. 2020. Physiological Response to Cycling With Variable Versus Constant Power Output. Frontiers in Physiology.
4. 04
   Atkinson & Brunskill 2000. Pacing strategies during a cycling time trial with simulated headwinds and tailwinds. Ergonomics.
5. 05
   Swain 1997. A model for optimizing cycling performance by varying power on hills and in wind. Medicine & Science in Sports & Exercise.
6. 06
   Seiler et al. 2013. Adaptations to aerobic interval training: interactive effects of exercise intensity and total work duration. Scandinavian Journal of Medicine & Science in Sports.

More inside How to do intervals outside

Other articles in this series

1. 01

   How long should a hill be for hill repeats?

   Rep length sets the hill. A table from 30-second to 10-minute repeats: how long, how steep, and which workout each hill serves.

   →
2. 02

   The best road for intervals: what to look for

   Long enough for the rep, no lights, a safe turnaround. How to find a stretch of road that lets an interval session run as written.

   →
3. 03

   VO2 max intervals outside: hill or flat road?

   Uphill and flat VO2 reps are not the same work. What the research says about each, and how to ride 5x4 when you have no hills.

   →
4. 04

   Zone 2 outside: how to plan a ride without stops

   Lights and junctions break an easy ride into pieces. How to plan a Zone 2 loop that stays steady, and what to do when you cannot.

   →