Stage 13 of the 2026 Vuelta a España runs 193.2km from Almuñécar to Loja on September 4, climbing 3,277m across three categorized tests before doing something this race hasn’t asked riders to do yet: finishing on a long, gradual come-down. The last categorized climb, Puerto de Granada, crests with 37km still to race. Everything after that is descent and false-flat, through olive country, into a finish that isn’t a summit and isn’t a bunch sprint. It isn’t a flat trap either.

That’s a genuinely different puzzle than the last several stages in this race. Stage 12 was decided at 2,151 meters, where Enric Mas attacked repeatedly on Calar Alto and only Oscar Onley could follow him to the line. Stage 10 hid its difficulty in a finish that only looked gentle. Stage 11 was a straightforward bunch sprint. Stage 13 doesn’t end on the thing that makes it hard. The hard part is over with more than an hour of racing left, and what riders do with that hour, not how they climbed Puerto de Granada, decides who’s still in position when Loja arrives.

Quick Verdict: Pacing a Stage That Ends With a Downhill

Section	Distance/Gradient	Role	Target Effort
Venta de la Cebada (cat. 1)	14.1km at 5.1%, km 31.2	Early selection point, way too soon to decide anything	78-85% FTP
Alto del Legionario (cat. 3)	15.4km at 3.1%, km 83.4	Long, shallow drag (a fatigue tax, not a selection climb)	68-76% FTP
Puerto de Granada (cat. 2)	7.4km at 5.6%, km 155.4	The last real climb, cresting 37km from the line	85-92% FTP
The descent and run-in	~37km, mostly downhill and false flat	Where position gets won or lost, not on the climb	Draft, don’t solo; ride by group dynamics
Final kilometers into Loja	Flat, then a shallow uphill drag late	Neither a sprint nor a summit finish	88-95% FTP, positioning matters more than watts

Race date: September 4, 2026 Stage distance: 193.2km, Almuñécar to Loja, 3,277m of elevation gain, classified a medium mountain stage The lesson: The climbing ends at km 155.4. The stage doesn’t end until km 193.2. Pace the 37km in between like it still matters, because it’s the part most riders will get wrong.

Why This Stage Breaks the Pattern

Every mountain stage we’ve covered in this Vuelta has shared one structural trait: the hardest terrain comes last. Aitana back-loaded its difficulty into a brutal final 22km. Calar Alto finished at altitude, on the summit, yesterday. Even the sprint stages saved their sting for the final kilometer. The unspoken rule of pacing this race has been simple: whatever’s left in the tank, save it for the end, because the end is where it gets hard.

Stage 13 inverts that. Puerto de Granada, 7.4km at 5.6%, is a real category 2 climb, steep enough in places to hurt, per La Vuelta’s official route page, and it arrives on legs that have already covered 155km and two prior climbs. But it crests with 37km still to go. What comes after isn’t recovery in the way a short descent between climbs is recovery. It’s the entire second act of the stage: a long downhill and a stretch of rolling false flat through the Vega’s olive groves. Then a finish in Loja that’s flat for most of its final kilometers before tipping up into a shallow drag in the closing few hundred meters, according to Cyclingstage.com’s stage 13 route guide. Riders who treat Puerto de Granada as the day’s test and coast the last 37km on autopilot are going to find out, late and expensively, that the stage wasn’t over when the climbing was.

The Pacing Problem Nobody’s Legs Are Trained For

Here’s what makes a long downhill run-in genuinely tricky, and it’s not really about power at all. Descending fast on tired legs is a bike-handling problem before it’s a fitness problem. Reaction times matter more than watts at 60-70kph, and the margin for a late brake or a bad line shrinks as fatigue and speed both climb. The fundamentals are consistent: stay smooth, look further down the road than feels natural, and don’t tighten your gap to the wheel in front of you on technical sections the way you would on a flat, straight run, since experienced descenders consistently open up following distance rather than close it once speeds get serious.

That last point is where the real pacing decision lives. A 37km descent-and-run-in isn’t won by the rider who descends fastest solo. It’s won, or at least not lost, by staying in a group that’s sharing the work, since drafting can save somewhere in the range of 27-40% of a rider’s energy output depending on speed, group size, and following distance, a number that gets more decisive, not less, on flat and false-flat terrain where nobody’s climbing away from anybody. A rider who attacks alone off Puerto de Granada to build a gap on the descent is betting solo watts against thirty-some kilometers of group efficiency. That bet rarely pays off unless the gap built on the climb itself was already substantial.

The Route, Broken Down

Venta de la Cebada (14.1km at 5.1%, cat. 1, km 31.2): The stage’s first real test, arriving early (barely 30km in, on a 193km day). It’s long enough and steep enough to matter, and per La Vuelta’s official stage page, it’s the first categorized climb of the day at a genuine 5.1% average over more than 14km. Nobody with 160km still to ride should be spending real matches here, no matter how strong the legs feel on kilometer 20.

Alto del Legionario (15.4km at 3.1%, cat. 3, km 83.4): The longest categorized climb of the day by distance, and also the gentlest by gradient. A 15.4km climb that barely clears 3% isn’t a selection point. It’s a long, grinding drag through the middle of the stage that taxes legs without ever forcing a hard decision. Riders who ride it patiently arrive at Puerto de Granada with something left. Riders who push it because “category 3” sounds minor find out 70km later that it wasn’t.

Puerto de Granada (7.4km at 5.6%, cat. 2, km 155.4): The last categorized climb, and the only one on this route that resembles the hard, selective tests this Vuelta has served up all race. Short by this race’s standards (7.4km), but a genuine 5.6% average after 155km in the legs is a real effort, and it’s the last point in the stage where a strong climber can put time into a weaker one by climbing alone. After this, the terrain stops rewarding pure climbing ability.

The descent and run-in (roughly 37km, Puerto de Granada’s summit to the finish): This is the stage, more than any single climb. A fast, technical-in-places descent gives way to gently rolling terrain through the Vega’s olive country, and the group dynamics here, who’s willing to work and whether a chase organizes, matter more than anybody’s individual FTP number.

The finish into Loja: The final kilometers flatten out almost completely before tipping into a shallow uphill drag in the closing stretch, per Cyclingstage.com’s route guide. That’s an awkward finish to read: not flat enough to guarantee a clean bunch sprint, not steep enough to be a pure climbers’ finish. Positioning into the final kilometer likely matters as much as raw sprint power.

How Do You Pace a Stage That Ends on a Downhill?

A finish decided by a long descent and run-in, rather than a climb, asks a different question than most mountain stages: not “how hard can I push,” but “how do I arrive at the front of the right group with the legs to finish it.” Follow this stage or face your own ride that ends with a long downhill run-in to a finish line, and the same rules apply:

1. Don’t burn the descent solo unless the gap is already real. A lone rider on a fast downhill is giving up 27-40% of the energy savings a group provides. Unless you built a meaningful advantage on the last climb, attacking into the descent usually just means getting caught by an organized chase with less in the tank to respond.
2. Open your following distance on technical sections, not just the straights. The instinct on a fast group descent is to stay glued to the wheel in front for the draft. That instinct is right on straight, sighted roads and wrong through corners. Give yourself room to react, because reaction time shrinks as speed climbs.
3. Treat false flat as work, not recovery. A 37km run-in reads as “downhill” on the elevation chart, but a meaningful chunk of it is rolling or false-flat, and false flat at pack speed after five hours of racing still costs real energy. Don’t coast through it mentally just because the profile looks friendly.
4. Save something specific for a finish that isn’t a sprint or a summit. A flat run-in into a late, shallow uphill drag rewards positioning into the last few hundred meters more than it rewards raw sprint wattage. Riders who spend their last matches on the descent, thinking the hard part is behind them, often have nothing left for that final kick.
5. Respect the first climb even though it’s 160km from the finish. Venta de la Cebada arrives at km 31, early enough that it’s tempting to ride it hard just because the legs are fresh. Everything that happens in the final hour of this stage depends on what’s left after 155km and three climbs, not what felt good at kilometer 30.

A Workout to Train This

You don’t need a mountain descent to train this pattern: a long, gradually descending or flat stretch of road, or a trainer session with a genuine no-recovery structure, teaches the same discipline. Climb with purpose, then hold a group-paced effort for longer than feels necessary, then finish with something left.

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

Block 1 (“Cebada”): 30-35 minutes at 78-85% FTP (RPE 5-6/10). Controlled, not casual: this is a real category 1 climb, just not the one that decides the day.

Block 2 (“Legionario”): 35-40 minutes at 68-76% FTP (RPE 4-5/10). The longest block in the session and the easiest to shortchange. Hold the number even when it feels like nothing is happening.

Block 3 (“Puerto de Granada”): 15-18 minutes at 85-92% FTP (RPE 7-8/10). The one genuinely hard effort of the session. Treat it like the last real test it is.

Block 4 (“the descent and run-in”): 25-28 minutes at 60-68% FTP (RPE 3-4/10), but don’t fully coast. Stay attentive: this block trains sustained, moderate output on tired legs, the same discipline a group-paced descent and false-flat run-in demands.

Block 5 (“the finish”): 8-10 minutes at 88-95% FTP (RPE 7-8/10), with the final 60-90 seconds as close to an all-out effort as you can manage.

Cool-down: 8-10 minutes easy.

Total time: 130-150 minutes.

The test: Compare your final 60-90 seconds in Block 5 to your best effort in Block 3. If Block 5’s finish holds up, you paced Block 4 correctly and protected something for the line. If it’s noticeably flat, that’s the exact mistake this stage punishes: spending too much on the last climb, or too much fighting the descent solo, and arriving at the finish with nothing left for a finish that demands one more effort. If your FTP number is more than six or eight weeks old, retest it before setting these targets, since a day this dependent on restraint after the climbing ends is a bad place to pace off a stale number.

What to Watch For Today

1. Whether Enric Mas’s team rides defensively after yesterday. Mas extended his advantage atop Calar Alto, with only Oscar Onley able to match his repeated attacks on the final climb. A medium mountain stage the day after a summit finish is a classic recovery-legs trap for the leader’s team: there’s little incentive to force the pace on Puerto de Granada, but also little margin to let a dangerous move ride away on the descent.
2. Whether the breakaway survives to Loja. The first 80km of this stage are hard enough, with Venta de la Cebada and the start of Legionario both inside that window, that keeping a strong break in check is difficult, and a nearly 200km stage with a long, workable run-in to the line is exactly the shape that lets an organized break stay clear.
3. How the group behaves on the descent off Puerto de Granada. This is the tell for the whole stage. A chase that organizes quickly and shares the work can close real time in 37km. A fractured, disorganized descent can let a small group build a gap that never comes back.
4. Who’s positioned well entering the final kilometer. With a flat run-in giving way to a late, shallow drag, this finish rewards teams that can deliver a rider to the front in the last 2-3km, closer to leadout tactics than a pure sprint or a pure climbers’ finish.

The Bottom Line

Stage 13 isn’t the hardest climbing day this Vuelta has produced, and it won’t be mistaken for one. What makes it worth pacing carefully is the shape: three climbs that get progressively less decisive as the stage goes, followed by 37km where the road points mostly downhill and the temptation is to treat that as the reward for the climbing that came before. It isn’t. It’s the part of the stage most likely to separate riders who read the whole profile from riders who stopped paying attention once the last categorized climb was behind them.

The transfer to your own riding doesn’t require a Spanish descent through olive country. Any ride that ends with a long downhill or a flat run-in after the climbing’s done, a century route or a group ride that tails off into rolling terrain, deserves the same respect this stage does. Draft instead of soloing, and open your gap on the technical bits instead of closing it. Save something specific for a finish line that isn’t a summit, because the climb ending doesn’t mean the stage did.

---

Stage distance, elevation gain, and climb data from La Vuelta’s official stage 13 page and Cyclingstage.com’s stage 13 route guide. Stage 12 result and GC context from ProCyclingUK’s stage 12 preview, referenced in our Stage 12 preview. Descending and drafting guidance from road.cc’s descending technique guide and TrainerRoad’s group riding and drafting guide. Training targets are estimates based on power-based training principles. Adjust for your own tested FTP and conditions.