Four days ago Tadej Pogacar rode up the Col du Tourmalet in 43 minutes and 12 seconds, the fastest ascent in Tour de France history. You will never hold 6.39 W/kg for three quarters of an hour. Neither will I. Here's the part that should cheer you up: most amateur riders lose far more time on climbs to bad pacing, wrong gears and wasted movement than to missing watts. This guide covers that skill layer, updated for July 2026 with the newest research, the current gearing options and a few lessons from the Tour that's happening right now.

Key takeaways

- Stay seated by default. Standing costs roughly 5–10% more oxygen at the same power. It only pays off above ~10% gradients, in short near-maximal surges, or as brief muscle relief.

- Pace by climb length. Roughly 115–130% of FTP for a 5-minute climb, 95–105% for 20 minutes, 90–100% for an hour. Start the first quarter below target.

- Spin 70–85 rpm seated. Chronic grinding under ~60 rpm is a gearing problem, not a toughness badge.

- Gear for your steepest local climb. Stock lowest ratios in 2026 run from 1.00 (Shimano 34x34) down to 0.81 (Campagnolo 45/29 with 11-36), and sub-compact setups are going mainstream.

- Train the skill. Four drills with exact protocols below: low-cadence torque work, over-unders, big-gear tempo and overgeared hill repeats.

You Don't Need Pogacar's Watts to Climb Better

Here's the uncomfortable math from Stage 6 of the 2026 Tour. Pogacar's Tourmalet record works out to an estimated 6.39 W/kg sustained for about 43 minutes, a VAM of roughly 1,761 vertical metres per hour. He then lifted it to about 7.2 W/kg over the final 15 minutes. A fit amateur puts out 2.5–3.5 W/kg on a good day. That gap is genetic and permanent. No winter training block closes it.

Now watch what actually happens on your club ride. The rider who gets dropped on the local 20-minute climb usually isn't the weakest engine in the group. It's the one who surged with the fast starters in the first three minutes, ran out of gears at the steep pitch, stood up at the wrong moment, and spent the last kilometre grinding at 55 rpm with their heart rate pinned. Not one of those is a fitness problem. They're skill errors, and skill errors are fixable this week rather than next season.

That's the scope of this article. We already cover engine-building (FTP work) elsewhere on the blog, so this post deliberately skips it. Instead we'll work through the four levers that decide how much of your existing fitness actually reaches the road on a climb:

1. Position: when to sit, when to stand, and what the research says each costs you.
2. Pacing: power and heart-rate targets by climb duration, and why surges are fatal.
3. Gearing: real 2026 options mapped to real gradients, with actual ratios and prices.
4. Drills: four protocols with sets, durations, cadences and intensities.

One self-check before we start, borrowed from Bicycling's May 2026 climbing guide: if you're breathing hard within the first 3 minutes of any climb, you started too fast. Ask yourself, "Is this a pace I can still live with 20 minutes from now?" If the honest answer is no, everything else in this article gets sabotaged before it has a chance to work.

What's New for Climbers in 2026

Climbing is having a moment. The 2026 Tour de France has already produced the fastest Tourmalet ascent ever: Pogacar's 43:12 for 17.1 km at 7.3% on 9 July beat the 45:35 benchmark he and Jonas Vingegaard set together in 2023 by more than two minutes. Vingegaard's 43:42 the same day would also have broken the old record. And the race has saved its biggest experiment for the finale, the first back-to-back Alpe d'Huez summit finishes in history on Stages 19 and 20. Stage 20 also crosses the Croix de Fer, the Télégraphe and the 2,642 m Col du Galibier, for roughly 5,600 m of climbing in a single day. Marco Pantani's 37:35 from 1997 remains the fastest recorded Tour ascent of Alpe d'Huez. For now.

The tech side is just as interesting, because for once the pro trends point in a rider-friendly direction: easier gears and shorter cranks.

• Sub-compact gearing has gone mainstream. 2026 test bikes are shipping with chainrings like 46/32 paired to an 11-39 cassette, a 0.82 low ratio that would have been dismissed as "touring gearing" five years ago.
• Cranks are getting shorter. Pogacar races on 165 mm cranks. A shorter crank lets you spin a higher cadence with a more open hip angle, which is exactly what seated climbing asks of you.
• Thirteen speeds are here. Campagnolo's Super Record 13 ships with the lowest stock climbing gearing of any flagship road group (numbers in the gearing section below), and most industry watchers expect a 13-speed wireless Shimano group next.

Why should a non-racer care? Because all of this filters down. The real message of 2026 is that low gears are no longer an admission of weakness. They're what the fastest riders in the world are choosing. If the winner of the Tour de France sees value in gearing that keeps his cadence high on a 7% road, the rider doing the same climb two and a half times slower needs that value far more.

Seated vs Standing: What the Research Actually Says

This is the most-argued question in climbing, and one of the few with genuinely good data behind it. The landmark study is Millet et al. 2002 (Medicine & Science in Sports & Exercise), which put trained cyclists on a simulated 6% climb and measured gross efficiency seated versus standing. At moderate intensity the results were blunt: 20.3% efficiency seated against 18.8% standing at 50% of peak power, and 20.7% against 19.4% at 70%. That's a 1.3–1.5 percentage-point penalty every time you leave the saddle.

The twist is what makes the study so useful. At 90% of peak power the penalty disappeared: 20.8% seated, 20.0% standing, no significant difference. In other words, standing stops costing you extra almost exactly when it becomes tactically useful. Near-maximal surges, attacks, steep-pitch efforts.

Ernst Hansen's research adds the gradient dimension. On slopes up to about 10%, sitting is more economical. Above 10%, standing lets you sustain higher power, at the cost of roughly 5% more oxygen. At a shallow 4% gradient ridden at 19 km/h, sitting needs about 10% less oxygen than standing for the same speed. A separate controlled uphill ergometer study found both higher VO2 and higher heart rate standing at identical power output. The physiology is consistent across every study: out of the saddle, you pay rent.

Here's the decision table that falls out of the research:

The transition matters as much as the position. The classic beginner mistake is standing up, momentarily stalling the bike and punching a hole in the group behind. The fix is mechanical. Shift up one gear before you stand (cadence naturally drops 10–20 rpm out of the saddle), rise as the pedal comes over the top of the stroke, and keep the bike tracking straight. Sit back down with the same smoothness, shifting back down a gear as you settle.

Pro tip: Count your standing time. Most riders think they stand for 20 seconds and actually stand for over a minute. If you're standing more than about 20% of a sub-10% climb, you're burning oxygen you'll want in the final kilometre.

Climbing Position: Small Fixes, Free Speed

You can gain real time on a climb without a single extra watt, just by letting the watts you have reach the pedals. Coaching sources are unusually aligned here. British Cycling and Roadman Cycling publish nearly identical seated-position cues, which for cycling technique is about as close to consensus as it gets.

Seated form checklist:

• Hands on the tops or hoods, not the drops. This opens your chest so your diaphragm can actually work. Aerodynamics barely matter at 10 km/h.
• Slide slightly back on the saddle to recruit your glutes, the biggest muscles you own. If only your quads are burning, you're sitting too far forward.
• Relax your shoulders and lighten your grip. A white-knuckle hold on the bars wastes energy in muscles that contribute nothing to forward motion.
• Keep your back straight and your knees tracking vertically. Knees flaring outward are a sign the gear is too big or the saddle is wrong.
• Shift up one gear before standing, so the natural cadence drop doesn't stall the bike.

Standing form is its own skill. Get your hips over the bottom bracket rather than thrown forward over the stem, and let the bike sway gently beneath you in rhythm with the pedal stroke. The sway should come from the bike, not from your torso. Don't pull the bars apart or wrestle the bike. On a climb, upper-body violence is just heat.

Breathing gets ignored because it feels automatic, but on a hard climb it's a pacing instrument. Deep, rhythmic, belly-first breathing delivers more air per breath than shallow chest panting, and it gives you a live gauge of effort with none of the lag of a head unit. This is where that 3-minute rule from Bicycling's 2026 guide earns its keep: breathing that turns ragged in the first 3 minutes is a red flag that you started above your sustainable pace. Back off now, not after you blow.

A practical scenario. You hit the base of a 25-minute climb feeling great, and your group's fast starters open a 50-metre gap inside two minutes. Position discipline (hands on tops, glutes loaded, breathing controlled, cadence up) plus the patience to let them go is very often how you ride past those same riders at minute eighteen. The climb doesn't reward whoever leads at the bottom.

How to Pace a Climb by Power and Heart Rate

Pacing is the highest-leverage skill in climbing. It's also the most quantified. The coaching consensus, compiled from Roadman Cycling's 2026 pacing guide and TrainRight's FTP testing work, gives clear targets by climb duration (FTP approximates your one-hour maximal steady power):

The single most valuable execution framework is the quarters strategy for long climbs:

1. First quarter: 90–95% of your target power, RPE 6–7. It should feel too easy. That's deliberate. Use power as a ceiling here.
2. Middle half: right at target, RPE 7–8. Settle, breathe, eat if it's a very long climb.
3. Final quarter: 105–110% of target, RPE 9–10. Now power becomes a floor. Don't let it sag as fatigue piles up. Empty the tank over the top.

Why is surging so catastrophic? The physiology has a name: W' (pronounced "W-prime"), your finite anaerobic work capacity. Every dig above FTP, whether it's chasing a wheel, punching over a steep ramp or showing off at the base, draws down a small battery that recharges slowly. Once W' is empty, even threshold power feels unsustainable and time-to-exhaustion collapses. The rider who climbs steady still has W' in reserve for the final quarter. The rider who surges spends it in the first ten minutes and pays compound interest the rest of the way up.

One more instrument warning: cardiac drift. On long, steady climbs your heart rate creeps up 5–15 bpm at constant power, thanks to heat and dehydration. So HR is a useful stabilisation check early in a climb, and it lies to you late in a climb. Hold a heart-rate number slavishly for an hour and you will be slowing down the whole time. The mature pacing stack looks like this: power (or the RPE table above) as primary, breathing as the live gauge, heart rate as an early-climb cross-check only.

For context on what the ceiling of human pacing looks like, Pogacar's Tourmalet record wasn't just big watts. It was negative-split execution: an estimated 6.39 W/kg overall, rising to about 7.2 W/kg over the final 15 minutes. Fastest at the end. That's the quarters strategy, executed at a level none of us will touch, and it's the same template that works at 3 W/kg.

Cadence: The Gradient-by-Gradient Guide

Flat-road cadence advice (80–95 rpm) doesn't survive contact with a mountain. Climbing cadence runs naturally lower, and the practical sweet spot for most amateur riders is 70–85 rpm seated. Below that, each pedal stroke becomes a mini leg-press: the load shifts from your cardiovascular system, which recovers in minutes, onto your muscles and knees, which don't.

Roadman Cycling's gradient-by-gradient guide is the cleanest reference we've found:

Two numbers in that table deserve a second look. First, standing cadence of 55–70 rpm is normal. The Indoor Cycling Institute defines standing-climb intervals at 65–85 rpm, and plenty of experienced riders settle near 55–60 rpm out of the saddle. Don't fight the drop. Plan for it — that's why you shift up a gear before standing. Second, the floor: chronic grinding below ~50–60 rpm on long climbs is a gearing problem, not a fitness badge. USA Cycling's coaching guidance is blunt about this. Mashing raises muscular strain and knee load, and the fix isn't more toughness. It's a bigger cassette or smaller chainrings.

A quick self-diagnostic. On your steepest regular climb, glance at your cadence at the hardest pitch while in your easiest gear. If you can't hold at least 65 rpm there, your gearing is the limiter, not your legs, and the next section is the most important one in this article for you. If you're comfortably above 80 rpm in your easiest gear with sprockets to spare, your gearing is fine and your gains live in the pacing and drills sections instead.

Climbing Gears for Real-World Gradients

Nobody in the pro peloton grinds up the Tourmalet over-geared anymore, and 2026 is the best year in history to stop doing it yourself. Here's what the big three groupset makers actually offer for climbing, with real ratios and prices:

The headline: Campagnolo's Super Record 13 currently ships the lowest stock climbing gear of the big three at 0.81. SRAM's 46/33 with a 10-36 gives you 0.92, and Shimano's road groups bottom out at exactly 1.00 with 34x34, pending the 13-speed wireless group the industry expects. For properly brutal terrain, SRAM's mullet and XPLR 1x13 routes go below 0.8, trading tight gear spacing for survival gearing.

How low do you need to go? A useful rule of thumb: a 1.00 ratio at 70 rpm moves you at roughly 8.8 km/h. Every step down in ratio lets you hold the same cadence at a lower speed, which is exactly what a steep gradient forces on you. Map it onto real climbs from this year's Tour route. The Col d'Aspin is 12 km at 6.5%, the Tourmalet 17.1 km at 7.3%. A strong amateur at 3 W/kg climbs a 7% gradient at roughly 9–11 km/h, which is fine on a 1.00 low gear. The moment the road kicks to 10–12%, that same rider slows to 6–7 km/h, and holding 70 rpm there demands a ratio in the 0.8s. This is why the sub-compact trend exists.

Gearing decision framework:

1. Find the steepest sustained gradient you actually ride. Not the steepest in your country, the steepest on your routes.
2. Ride it in your current easiest gear and note your cadence at the hardest pitch.
3. 65+ rpm? Your gearing is adequate. Spend your money elsewhere.
4. Below 65 rpm? Gear down one step: an 11-34 cassette first (cheapest), then sub-compact or 46/33-style chainrings, then the Campagnolo, mullet or XPLR route for true alpine or loaded riding.
5. Re-test. Gearing is solved when the answer to step 3 is yes on every climb you care about.

For a deeper dive into how the three brands' groupsets compare beyond climbing gears, see our Shimano 105 vs SRAM Rival vs Campagnolo comparison.

Four Climbing Drills That Actually Work

"Do hill repeats" is not a training plan. These four sessions come from named coaching sources with exact protocols: sets, durations, cadences and intensities. Pick two per week, never on back-to-back days, and treat them as quality sessions with easy riding around them.

1. Low-cadence torque work (CTS/TrainRight). The classic climbing-strength session: 4x5 minutes at 50–55 rpm, at tempo-to-sweet-spot power (80–95% of FTP), with equal-duration easy-spin recoveries. Ideally done on a hill around 6%. Build total work time from about 20 minutes as a beginner toward 40 minutes over several weeks. This teaches your legs to produce climbing torque without your heart rate exploding.

2. Over-unders (coach John Wakefield, Science to Sport). The "don't get dropped when the pace lifts" session: 3x10 minutes alternating 2 minutes at 95% FTP with 1 minute at 110% FTP, cadence 85–95 rpm, 10 minutes easy between sets, on flat road or a steady 3–8% gradient. This trains exactly the skill group climbs punish: absorbing surges above threshold and recovering while still riding hard. In other words, managing W' under pressure.

3. Big-gear tempo (Road Cycling Academy). A gentler torque alternative: 5x5 minutes at around 60 rpm in Zone 3 (76–90% FTP) with 3–4 minutes of easy, high-cadence spinning between efforts. One hard rule: don't go below ~50 rpm. Beneath that, knee load rises faster than the training benefit.

4. Overgeared hill repeats (USA Cycling). The simplest of the four: 3x8 minutes climbing in a gear that cuts your preferred cadence by at least 10 rpm. If you'd normally climb this hill at 80 rpm, ride it at 70 or below. Full recovery between reps.

No hills? No excuse. All four sessions adapt to flat terrain. Low-cadence and big-gear work translate directly to an indoor trainer with the resistance cranked up, or a steady headwind stretch. Over-unders were designed to work on flat roads in the first place. Overgeared repeats work anywhere you can hold 8 minutes of continuous effort, including the much-mocked but effective parking-garage ramp loop. Flatlanders who train torque and threshold surges routinely out-climb mountain-dwellers who just ride their hills easy.

Fueling and the Mental Game on Long Climbs

The strongest legs of your life are useless if they're empty at the base of the final climb. USA Cycling's coaching guidance is specific: eat 30–60 g of carbohydrate per hour on long rides, and top up 15–20 minutes before a major climb. A gel or half a bar, timed so the sugar arrives in your blood as the road tilts up. Don't try to eat mid-climb on anything steep. Chewing at threshold heart rate is a skill nobody needs. Eat on the flats and descents before the climb instead. (For the full picture, see our cycling nutrition guide for long rides.)

Then there's the head. Two mental techniques show up in essentially every credible climbing guide, including Bicycling's 2026 update:

• Chunk the climb. Never ride to the summit. Ride to the next switchback, the next road sign, the next shadow line. A 60-minute climb is unrideable as a single thought and trivial as forty 90-second segments. Pogacar doesn't ride 17.1 km of Tourmalet; he rides the next 500 metres, repeatedly.
• Climb at your pace, not the group's. The single most common amateur blow-up is riding the first ten minutes at someone else's sustainable power. Let the gap open. On any climb over 15 minutes, steady riders reel back surgers with brutal reliability. That isn't motivation talk. It's the W' physiology from the pacing section.

Pre-climb checklist (run it in the last 2 km before the base):

• [ ] Eaten in the last 40 minutes? Gel now if not (15–20 min lead time).
• [ ] Drunk in the last 15 minutes? Two big sips. Cardiac drift punishes dehydration.
• [ ] Shifted into the small chainring before the gradient bites, not under full load on the ramp.
• [ ] Target power/RPE chosen from the pacing table for this climb's length?
• [ ] Zipper, helmet vents, sleeves. Sort your heat management while your hands are still free.

One equipment note that costs nothing to get right: climbing days are hot days working hard. A lightweight jersey that breathes properly, with deep pockets for the gels and the stowed vest, is what makes the fueling plan above workable at 40 heartbeats below your max. That's precisely the use case our Rydecruz climbing jerseys are cut for.

FAQ: Climbing Questions Riders Actually Ask

Q: Should I sit or stand when climbing on a road bike? A: Sit by default. Research (Millet 2002) shows standing is 1.3–1.5 efficiency points worse at moderate intensity and costs roughly 5–10% more oxygen for the same power. Stand when the gradient goes past ~10%, for short surges near maximal effort (where the efficiency gap disappears), or for 15–30 seconds every few minutes to give your muscles a break.

Q: What is the best cadence for climbing hills? A: 70–85 rpm seated for most amateur riders, which is lower than flat-road cadence. Scale it to gradient: 80–90 rpm at 3–5%, 75–85 rpm at 6–8%, 70–80 rpm at 9–12%, 65–75 rpm above 12%. Standing, expect your cadence to drop 10–20 rpm to around 55–70.

Q: Why do I get dropped on climbs even though I'm fine on the flats? A: Climbing exposes watts per kilogram, pacing and gearing rather than raw watts. The usual culprits are starting the climb above your sustainable power, which drains W' (your finite anaerobic battery), and gearing that forces your cadence below ~65 rpm on steep pitches. Fix the pacing with the quarters strategy and the gearing with a bigger cassette before you assume you need a bigger engine.

Q: What gears do I need for steep climbs of 10% or more? A: Aim for a lowest ratio near or below 1.0, and into the 0.8s for sustained 10%+ gradients. Stock 2026 options: Shimano 34x34 (1.00), SRAM 46/33 with 10-36 (0.92), Campagnolo Super Record 13 45/29 with 11-36 (0.81), or sub-compact 46/32 with 11-39 (0.82). SRAM mullet (10-52) and XPLR 1x13 builds go below 0.8 for the steepest terrain.

Q: How do I pace a long climb so I don't blow up halfway? A: Use the quarters strategy: first quarter at 90–95% of your target power (it should feel too easy), middle half at target, final quarter at 105–110%. Targets by duration: roughly 95–105% of FTP for a 20-minute climb, 90–100% for an hour. And trust breathing over heart rate late in the climb, because cardiac drift adds 5–15 bpm at constant power.

Q: Is it bad to grind uphill at a really low cadence? A: Chronically, yes. Below ~50–60 rpm the load shifts from your cardiovascular system onto muscles and knees, raising strain and injury risk. As a drill (say, 4x5 min at 50–55 rpm at 80–95% FTP), low cadence builds climbing strength. As your everyday climbing style, it means your gearing is too big.

Q: How can I train for long climbs if I live somewhere flat? A: Train the demands, not the scenery. Low-cadence torque work (4x5 min at 50–55 rpm) and big-gear tempo on an indoor trainer or into a headwind replicate climbing's muscular load, and over-unders (2 min at 95% / 1 min at 110% FTP) build surge tolerance on dead-flat roads. The physiology transfers almost completely.

Q: How many watts per kilogram do the pros put out on climbs? A: Pogacar's July 2026 Tourmalet record was ridden at an estimated 6.39 W/kg for about 43 minutes, rising to roughly 7.2 W/kg in the final 15 minutes. Strong amateurs typically sustain 2.5–3.5 W/kg on long climbs, which is exactly why pacing, gearing and technique matter more for the rest of us than for him.

The Bottom Line: Your Next Climb, Not Next Season

Everything above compresses into one ride plan. On your very next climb: pick your power or RPE target from the pacing table before the road tilts up, start the first quarter feeling suspiciously easy, hold 70–85 rpm seated with your hands on the tops, stand only for the 10%+ pitches and short surges (shift up a gear first), and empty whatever's left over the final quarter. This week, run the gearing diagnostic. If you can't spin 65 rpm on your steepest local pitch, an 11-34 cassette is the cheapest speed you will ever buy. Then slot two of the four drills into your training week and let them compound.

Pogacar's 43:12 on the Tourmalet was set with 6.39 W/kg that nobody reading this owns. But the negative split, the seated discipline, the 165 mm cranks spinning an easy gear, the surge control — that part of the record is a free download. The watts are his. The skills are yours.