South Downs Way 100 training: the runnable hundred that punishes bad pacing

The South Downs Way 100, run by Centurion, is 100 miles along the chalk of the South Downs, Matterley Bowl near Winchester to the athletics track at Eastbourne, with 3,800 m of ascent and a 30-hour limit. It's 85% runnable trail, and that word is the whole point. This is not a race the terrain slows down for you.

Runnable is not the same as easy

On a mountain hundred, the climbs force you to walk and the technical ground caps your speed. The South Downs give you none of that mercy. The gradients are gentle and almost everything is runnable, which means the pace is set by your discipline, not the ground. Go out at a pace that feels comfortable at mile 20 and the chalk will quietly break you by mile 70. With nowhere to hide, the SDW rewards even pacing and honest aerobic fitness more than any mountain race does, and it exposes overreaching without pity.

A hot Saturday before Washington

Centurion doesn't publish retirements by aid station, but its own race reports say where the damage lands. In 2023 the heat held from Saturday morning into the late evening. Of 339 starters, 192 finished: 57%, the lowest finish rate the race has had. The report saw leading runners struggling with the heat at Houghton Farm, mile 44, and at Washington, about mile 55. Then: "But when the sun did go down, the drop out rate fell off of a cliff". The 2016 report made the pacing point: "those who pushed hard came unstuck and suffered through the middle portion of the race." In milder years far more finish: 405 of 532 starters in 2025. So the danger is the first half on a hot June Saturday, at a pace that feels easy.

The cap earns its keep here

Because the whole route is runnable, holding a genuinely easy effort early is the entire skill. A heart-rate cap is the cleanest way to enforce it: it removes the negotiation, it sees the heat and the fatigue a pace target misses, and it keeps the first 50 miles cheap so there's something left for the second. On a runnable course, the early miles are where a runner spends too much because it felt fine. The easy-running field note is the method; this race is where the discipline pays out most obviously.

Fuel the pace you're actually holding

Because you're running, not hiking, you're burning at a higher rate for longer, and the fuelling has to match. A target of grams per hour you can absorb while running, started early and held, tested on long runs at race effort. The daytime June heat on open downland adds a fluid and sodium demand on top. Drink to thirst, never treat sodium as licence to overdrink, and rehearse the whole plan before the day.

Where durability shows up

The SDW is the purest durability test in this library, because the terrain removes every excuse. Durability here is simply the pace you can still hold in the final quarter at a controlled effort, and it reads cleanly: on runnable ground the late-run efficiency numbers I use are at their most reliable. What I watch through a build is exactly that, the heart-rate cost of holding pace late in long runs, and whether that cost is improving across comparable sessions. This is the race where the data is most trustworthy and most useful. It still informs a coaching decision rather than making one.

Twenty-four weeks to Matterley Bowl

For the 12 June 2027 start I build over 24 weeks from late December. Capped easy volume is most of the work for a runnable hundred, with two strength sessions a week to protect the legs. Long runs step up by about 10% and hold after three weeks of progress.

• Capped long runs with a closing check: pace in the last hour against the first, at the same heart rate. If it falls away, the start was too quick.
• A 50-mile race about eight weeks out, run at 100-mile effort, in place of one giant training run.
• Running fuel: a gut ladder from about 30 g of carbohydrate an hour, adding 10 g every one to two weeks, settled two weeks out.
• Back-to-backs on the chalk every two to three weeks in the Downs block, about five hours then about four.
• Hot-day long runs when May offers them, still under the cap, with fluid and sodium rehearsed.

The biggest back-to-back lands about five weeks out. Some gentle downhill volume stays in throughout, because 3,800 m of runnable descent still bills the quads. Night practice too, because most of the field runs through one. Then a three-week taper that banks the fitness and spends the fatigue.

What I watch

• The heart-rate cost of holding pace late in long runs, against the early miles.
• Whether that late-run cost improves across comparable sessions.
• Pacing discipline: how easy the early miles actually are, not how easy they feel.
• Fuelling at running effort: grams per hour logged, gut score out of ten.
• Quad condition late in long runs, because runnable descent still accumulates.
• Recovery and heat response as the long runs peak into summer.

Facts checked September 2026 against Centurion's published information: 100 miles, 3,800 m of ascent, a 30-hour limit, starting at Matterley Bowl and finishing on the Eastbourne track, 12 aid stations, pacers allowed from mile 50, a 50-mile sub-15-hour qualifier. The 2025 edition finished around 76%. The route does not cross the Seven Sisters; it turns inland via Jevington. Re-checked October 2026: the 2027 race page gives Saturday 12 June 2027, two start waves, and Washington at mile 55 on its aid-station table. Where runners stop: Centurion publishes no breakdown by aid station, so this comes from its own race reports, read October 2026: 2023 (339 starters, 192 finishers, the heat, Houghton Farm and Washington, the quote on sunset), 2016 (the quote on pacing) and 2025 (532 starters, 405 finishers). The build numbers are my coaching method: a long-run step of about 10% with a hold after three progress weeks, a gut ladder in 10 g steps finished two weeks out, a second back-to-back day at about 80% of the first and a three-week taper. Evidence level: the physiology matches the field notes; the pacing argument is coaching practice the runnable terrain makes unusually clear.