Two days into this Vuelta and the GC picture is already tight enough to lose sleep over. Tadej Pogačar beat Ethan Hayter by 0.09 seconds in Monaco, as close as a time trial result gets without a photo finish argument, and stage 2 kept the bunch together for a bunch sprint. So nobody’s actually lost anything yet. That changes tomorrow. Stage 3 of the 2026 Vuelta a España runs 174km from Gruissan-Aude to Font Romeu, and it’s the first summit finish of the race (one of seven on this route), the first place GC time actually starts moving.

We paced Monaco’s technical time trial a few days ago and called it a handling problem wearing a power-test costume. Font Romeu is the opposite kind of puzzle, and a trickier one than it looks on paper. It isn’t one climb. It’s the category 1 Col de Mont-Louis run straight into the category 2 finishing ascent, with barely any flat road stitching them together. Pace that like a single effort and you’ll find out the hard way that it isn’t one.

Quick Verdict: Pacing a Stacked Climb Finish

Detail	Info
Distance	174km, Gruissan-Aude to Font Romeu
First climb	Col de Mont-Louis, cat. 1: 19.4km at 4.8%
Finishing climb	Font Romeu, cat. 2: 9.8km at ~5%, summit at 1,936m
Gap between climbs	Minimal, effectively one long effort split into two distinct gradients
What it rewards	Restraint on the first climb, a second gear for the second
Amateur mistake	Treating the cat. 1 climb as the main event and arriving at Font Romeu empty
Stakes	First of seven summit finishes, the first real GC sorting point of the race

The lesson: A stacked finish isn’t one hard climb with a warm-up. It’s two separate pacing decisions, and the first one determines whether you get to make the second.

Why Two Climbs Beat One Climb of the Same Length

Add up Mont-Louis and Font Romeu and you get something close to 29 kilometers of net climbing before the finish line. That’s longer than the hour-plus Ventoux grind we wrote about earlier this month, and on paper it sounds like the same problem: one long sustained effort, pace it evenly, don’t blow up early. It isn’t the same problem. Ventoux is a single gradient that barely changes shape for 15.7km. Font Romeu’s finale is two climbs with two different personalities glued together, and the transition between them is where riders who paced it wrong find out.

Mont-Louis is the deceptive one. At 4.8% average over 19.4km, it’s long but not steep, the kind of gradient that lets a strong rider sit at 90% FTP for the better part of 45 minutes and feel like they’re cruising. That feeling is the trap. Font Romeu’s finishing climb starts almost the moment Mont-Louis ends, and it’s steeper on average (roughly 5% over 9.8km) with pitches that bite harder than the number suggests, especially in the closing kilometers into a finish sitting at 1,936m. A rider who spent Mont-Louis riding at the edge of what’s sustainable doesn’t have a second gear left for that. They have whatever’s left, which on a long, deceptively gentle first climb is usually less than they think.

The Two-Climb Problem, Broken Down

Think of this finale as two separate pacing decisions stacked on top of each other, not one long one.

The Col de Mont-Louis (19.4km at 4.8%): This is a discipline test, not a power test. The gradient is gentle enough that going 5-8% over your sustainable number won’t hurt immediately. It’ll hurt in 25 minutes, on a different climb, when you can’t afford it. Ride this section at 82-88% FTP. That’s conservative for a category 1 climb, and it’s supposed to feel that way.

The transition: There’s little to no flat road between the two ascents. Don’t treat this as a recovery zone. There isn’t one. The best you can do is settle your breathing and resist the urge to answer any attacks that go on the lower slopes of Font Romeu, where the road first kicks up again.

Font Romeu itself (9.8km at ~5%, finishing at 1,936m): This is where the stage gets decided, and it’s where your legs need something in reserve. Target 88-95% FTP here, building toward the finish rather than starting hot. A rider who paced Mont-Louis correctly should be able to ride Font Romeu harder than Mont-Louis — that’s the entire test.

How Should You Pace a Stacked Climb Finish?

Any time an event stacks two climbs back-to-back with little recovery between them (this stage, a hilly sportive, a local ride with a long drag into a steep finish) the same four decisions apply:

1. Ride the first climb slower than it feels like you need to. If the gradient is gentle enough that a hard pace doesn’t hurt yet, that’s exactly why it’s dangerous. Save the discomfort for the climb that actually decides the day.
2. Don’t treat the gap between climbs as recovery: treat it as a warning. A short or nonexistent transition means the second climb starts on legs that are already working. Settle, don’t surge, and don’t respond to early moves.
3. Reserve your hardest effort for the steeper, later gradient. The climb with less distance left to survive is where spent matches cost the most. Ride the harder climb harder, not the easier one.
4. Pace by what’s left, not by what’s already spent. Riders who blow up on stacked finishes almost always paced the first climb against their fitness rather than against what the second climb would demand. Budget backward from the finish line, not forward from the start.

A Workout to Train This

You don’t need a Pyrenean double-climb to train this pattern. Any ride with a long, gentle drag followed by a shorter, steeper pitch works, or you can build the shape on a trainer.

Warm-up: 15 minutes progressive, ending with two 30-second openers at 100% FTP.

The “Mont-Louis” block:

1. 35-40 minutes at 82-88% FTP (RPE 5-6/10): long, controlled, deliberately restrained. If this feels too easy for the first 20 minutes, that’s correct.

The transition: 2-3 minutes at 75% FTP. No coasting, no full recovery. Simulates riding through rather than resting.

The “Font Romeu” block:

1. 15-18 minutes at 88-95% FTP (RPE 7-8/10), building through the effort rather than starting at the top of the range.
2. Final 2-3 minutes at 100-105% FTP (RPE 9/10): the closing pitch into the finish.

Cool-down: 10-15 minutes easy.

Total time: 75-90 minutes.

The test: Compare your power in the final 2-3 minutes to the number you held through the Mont-Louis block. If it’s higher, or at least equal, you paced the long climb correctly. If it collapses, you spent too much on the gentle section and paid for it on the steep one. That’s the exact mistake this stage is built to punish. Run this session on any long, two-stage climb near you, or split a single sustained hill into a “conservative first two-thirds, harder final third” pattern if you don’t have two distinct gradients to work with.

If your FTP number is more than six or eight weeks old, retest it before setting targets off these percentages. A stale number on a two-stage effort like this one is worse than useless. It’ll tell you Mont-Louis is manageable right up until Font Romeu proves it wasn’t.

What to Watch For Tomorrow

1. Who attacks on Mont-Louis, not just on Font Romeu. A long, gentle cat. 1 climb tempts overconfident riders into early moves. Watch whether those moves survive the transition into the steeper finish.
2. The gap at the base of Font Romeu. If the front group is still large when the road kicks up for the finish, the real selection happens in the final 9.8km — a short, violent version of the GC test this whole route is built around, the one we mapped out in the full route breakdown.
3. Whether Pogačar’s 0.09-second cushion survives the day. His margin over Hayter from Monaco is small enough that a single bad kilometer on a summit finish erases it entirely. This is the first climb where that’s actually possible.
4. How Primož Roglič responds on the steeper gradient. He’s riding back from a training crash with a compressed build-up, and a summit finish after 174km is a much harder question than a short technical TT was.

The Bottom Line

Font Romeu isn’t one hard climb. It’s two climbs of different characters stitched together with almost nothing between them, and the entire pacing problem comes down to treating them as separate decisions rather than one long grind. Ride the gentle Col de Mont-Louis like it’s actually gentle (because it is, and it’s designed to fool you into thinking otherwise), and save the harder number for the steeper 9.8km that actually decides the stage.

For amateurs, the transfer is direct: any ride or event that stacks a long, easy-feeling climb into a shorter, steeper one rewards exactly this discipline. Pace the first section against what the second one demands, not against how the first one feels. That’s the whole lesson, and it’s the same one the GC contenders will either learn or get punished for tomorrow afternoon.

---

Stage distance, climb lengths, gradients, and summit altitude from La Vuelta’s official stage 3 page. Stage 1 result from CyclingUpToDate’s stage 1 report. Training targets are estimates based on power-based training principles. Adjust for your own tested FTP and conditions.