Brandon McNulty won the men's road race at the UCI Road World Championships in Montreal on 27 September 2026, after 273.4 km and 3,803 m of climbing. Velo's analysis shows his last climb of the Voie Camillien-Houde took 5 minutes 9 seconds, against 3 minutes 40 earlier, and he still won by 13 seconds, because his chasers had faded too.
What happened in Montreal
The UCI's race report sets out the course. 184 riders from 61 nations rode a loop of around 100 km before 12 laps of a 13.4 km circuit on Mont-Royal, each with the Voie Camillien-Houde, 1.67 km at 7.4%. McNulty attacked with 33 km to go and held on to finish 13 seconds ahead of Michael Matthews, with Mathieu van der Poel third.
Zach Nehr, a professional rider and coach, analysed the riders' power files for Velo, and the race numbers below are his, as Velo reported them. WattPlan hasn't checked them against the original files.
What the power files show
| Rider | Part of the race | Average power | Time |
|---|---|---|---|
| Quinn Simmons | First 100 km | 204 W | not reported |
| Quinn Simmons | Final 173 km | 346 W | not reported |
| Quinn Simmons | Whole race | 298 W | 6:17:23 |
| Bruno Armirail | Long pull on the circuit | 406 W | 1:32:27 |
| Brandon McNulty | Camillien-Houde, about 80 km to go | about 8 W/kg | 3:40 |
| Brandon McNulty | Camillien-Houde, two laps to go | about 6 W/kg | 4:36 |
| Brandon McNulty | Camillien-Houde, last lap | about 5.3 W/kg | 5:09 |
Quinn Simmons, who finished fourth, rode the opening 100 km at 59% of the power he averaged for the rest of the race. Armirail, riding for France, pulled the bunch at more than 400 W for an hour and a half to close down the early break, and Nehr notes that France's leader, Paul Seixas, later cramped out of the chase group. McNulty's lap times after his attack were 18:44, 19:39 and 20:16, so he slowed on every lap. In Nehr's reading every rider slowed over the closing laps, and no one was setting personal bests on the last one.
What is cycling durability?
Ed Maunder and colleagues gave the idea its working definition in a 2021 review in Sports Medicine. Durability is when, and by how much, your physiology deteriorates during prolonged exercise. Your FTP, thresholds and power curve are measured fresh, and they drift once you have been riding for hours.
How much you lose depends on how the earlier work was done. A 2025 systematic review by Sánchez-Jiménez and colleagues pooled 21 studies of cyclists and found that hard efforts, above critical power, cut later power more than the same amount of work ridden at a lower intensity. Declines of 10 to 20% followed 2.5 to 15 kJ per kg of prior high-intensity work. The authors add that kilojoules alone are a weak measure of fatigue, because they ignore how hard the work was.
Does durability matter for amateurs?
The evidence says yes, from small studies. Barsumyan and colleagues tested 14 trained amateurs in 2025, average age 37.5, training about 9.6 hours a week. Each did a 5 minute and a 20 minute time trial fresh, then again after riding at 70 to 80% of their fresh 20 minute power until they had done 1,000 kJ of work. Seven had stood on a podium in domestic races or qualified for a UCI world championship the season before, and seven raced without those results.
The successful group's 20 minute power fell 6.5% after the 1,000 kJ, half the 12.5% drop in the less successful group. Their 5 minute power and heart rate responses didn't differ. At 200 W, 1,000 kJ takes about 83 minutes, so on a sportive that fatigue arrives in the second hour, long before the climbs that usually decide the day.
Our view at WattPlan is that for a sportive or gravel event longer than three hours, the power you can still produce in the final hour predicts your day better than your FTP, and the cheapest gains come from a slower start and earlier fuelling before any extra intervals.
How should you pace yourself on a long bike ride?
Ride the first hours well below what feels sustainable. Simmons is a professional in a sheltered bunch, so 204 W cost him very little, while an amateur on a sportive rides far more of the day in the wind. The review found that efforts above critical power spend your durability fastest, so every surge to hold a fast group in the first hour is paid for on the last climb.
WattPlan's Route analysis works this way for any ride over an hour. Upload the event GPX and it sets each segment's target from the power you can sustain for the full predicted time, using your durability profile, warns you to avoid surges when W' falls below 40%, and exports a pacing GPX for your bike computer. Our study of 6,589 Quebrantahuesos finishers shows the same shape in amateurs, with the fastest riders starting slower.
Does fuelling change how well your power holds?
In the lab, yes. Norte and colleagues had 16 trained cyclists and triathletes, 15 men and one woman, ride 3 hours at 95% of their gas exchange threshold three times, drinking water only, 60 g of carbohydrate an hour or 120 g an hour, taken as a drink every 20 minutes. Fresh, their critical power averaged 277 W. After 3 hours it was 236 W on water, 257 W at 60 g and 266 W at 120 g, so the more carbohydrate they took in, the less they lost. The trial was randomised, the water ride had no placebo drink, a sports nutrition company supplied the carbohydrate and one author works for another.
WattPlan's in-ride reminders prompt 60 g of carbohydrate every 45 minutes, 80 g an hour, on rides of an hour or more.
How do you train cycling durability?
Put efforts late in long rides, not only into fresh interval sessions. A long ride with 2 or 3 blocks of 10 to 20 minutes at the power you expect to hold on the event's last climb, ridden in its final third, rehearses the part of the day the files show is decisive. Fuel those rides as you will on the day, because the Norte trial suggests fuelling is part of what you're measuring.
When you give a WattPlan plan your event and its expected time, its long rides build towards between half and 70% of that time, within the time you have on your long-ride day. The late efforts are yours to add.
How do you know you're ready for a long event?
Compare your power late in long rides with your fresh power. Ride your hardest 20 minutes in the last hour of a 4 hour ride and set it against your fresh 20 minute power, then repeat the same ride, fuelled the same way, every few weeks. The trend is what matters, and a gap that shrinks means your durability is improving. Don't hold your number against a study's. Barsumyan's successful amateurs lost 6.5% and the others 12.5%, but that was after a controlled lab protocol of 1,000 kJ at set intensities, a 5 minute test, 40 minutes of recovery and standard fuelling, so a test on the road isn't comparable.
WattPlan's Durability Score estimates your power at 3, 5 and 7 hours and scores you from 0 to 100 on your predicted 5 hour power, where 60% of FTP scores 0 and 85% scores 100. It now learns from your own long rides, comparing the average power in the second half of each ride with the first. Durability Score and Route analysis are part of Premium.
What to do before your next long event
- Set a power ceiling for the first two hours that feels too easy, and hold it when the group surges.
- Start fuelling in the first hour and keep it going every 20 to 45 minutes. The Norte riders lost less power at 120 g an hour than at 60 g, so build up the amount you can take in training.
- Once a week, finish your long ride with 2 or 3 blocks at your expected climbing power in its final third.
- Judge your readiness by the power you hold in the last hour of your longest rides, not by your FTP.
- Upload the course to Route analysis and hold its targets when the start gets fast.
What we don't know yet
The Montreal numbers are Velo's reading of three American files and Armirail's, and nobody has published the riders' weights, so the W/kg figures are approximate. They describe professionals racing in a bunch on one course, at a fitness no amateur shares.
The amateur evidence is small. Barsumyan's study had 14 riders, fewer than the 26 its own power calculation called for, mostly men, and it shows an association rather than proof that training durability improves results. The Norte trial was 16 riders in a lab, fully carbohydrate loaded beforehand. Researchers also measure durability in different ways, kilojoules, prior efforts or time, which makes their results hard to compare, and no trial yet shows which training improves it most in amateurs.
Sources
- 2026 UCI Road World Championships: McNulty caps off memorable Montréal week, UCI
- Power Analysis: How Brandon McNulty Won Worlds, Velo
- The Importance of 'Durability' in the Physiological Profiling of Endurance Athletes, Sports Medicine
- Is intensity the most important factor in determining the amount of prior work accumulated that affects cyclists' acute durability? A systematic review, European Journal of Applied Physiology
- Enhanced durability predicts success in amateur road cycling: evidence of power output declines, Frontiers in Sports and Active Living
- Graded Carbohydrate Ingestion up to 120 g·h(-1) Attenuates the Reduction in Critical Power Following 3 h of Moderate-Intensity Exercise in a Dose-Dependent Manner, Scandinavian Journal of Medicine and Science in Sports