Race to the Stones training: 100 km on the Ridgeway, often as a first ultra

Race to the Stones runs 100 km along the Ridgeway, from Lewknor in the Chilterns to the stone circle at Avebury. The 2027 race is on 10 and 11 July. The climbing is modest, 1,350 metres in total, and the organiser says you don't need technical skill. Red arrows mark the route and reflect at night. Seven pit stops sit along it, with water points between them. The organiser puts the completion rate at over 90 per cent.

That's why so many people pick it as their first ultra, and I'm happy when they do. The same features explain how people come unstuck. The course won't slow you down in the first 30 km, so your plan has to.

One push or two days

There's no walking category. You choose a format instead: the 100K non-stop, the 100K over two days with a night at basecamp on Lattin Down, or one of the two 50Ks. The two-day format splits the climbing almost evenly, 720 metres on day one and 630 on day two, with 15+ hours allowed each day. Non-stop runners get the longer window, but the organiser's pages disagree on how long. The route page says 30+ hours. The event page says 25+. I plan to the shorter figure until Threshold confirms which applies.

For a marathon runner over 40, I start by asking what the first finish is for. Two days gives you a 50K you recover from overnight, then a second morning that teaches you how legs behave when they start tired. The single push needs a longer build and a plan for the night. Bag rules differ too: two-day entrants get 10 kg, non-stop runners get 5 kg and can't reach it at basecamp. The first-ultra note covers the step from 42 to 50 km. The Stones doubles that, so I want the 50K behind you first.

Forty per cent walked

The organiser's own figure for 2025 is that 40 per cent of the field hiked or mixed walking and jogging. I read that as permission: a first 100K that's half walked is a normal 100K. What I want is walking planned from the first hour: short walks on every climb from the start, at a heart rate you could hold all day. The alternative is forced walking after 60 km, when the legs make the decision for you.

The runnable trap

The organiser describes long stretches of flowing, runnable trail, and that invites a fast start. Runner reports describe the same mistakes: going off too fast because it feels easy, dropping the fuelling routine once thinking goes fuzzy, and boredom on the long runnable sections.

So I set the first 30 km by heart rate, under the cap from the easy-runs note. Fuelling goes on a timer, because a timer keeps working after your judgement stops. Starts go in waves by predicted finish time; in 2025 they ran from 06:30 to 08:50. Give an honest prediction.

Where the non-stop field stops

Threshold publishes non-stop results with times at pit stop 2, at halfway and at pit stop 6. I counted the runners listed without a finish. In 2023, 874 of 938 finished. In 2024 it was 1,037 of 1,234, and in 2025, 715 of 812. That's 84 to 93 per cent, a little under the organiser's "over 90 per cent", which may cover every format.

Where people stopped changed from year to year. In 2024, 166 of the 197 non-finishers had a halfway time: they reached basecamp, then stopped there or later. In 2025, 84 of the 97 had no time past pit stop 2, about 21 km in. The results don't say why. A warm day, a fast start or a camp full of two-day runners settling in for the night could each play a part. So I plan for two places to fail: the first half on a warm day, and the walk out of basecamp.

The second half has its own cost. One runner on the 2026 second-day 50K found the trail "a lot more rutted and bumpy, making maintaining a decent running rhythm on tired legs much harder." Their feet were sore by then too, and a sock change at a pit stop helped.

July on the Ridgeway

The race sits in mid-July, and the organiser publishes its own advice on staying cool. One runner's report from the 2026 Ridgeline 50K describes the day warming from 18°C to 31°C, with the early miles banked in the cool and more walking than planned later on. That's a single runner's day. It's still enough for me to put a heat block into the last three weeks of a Stones build, using the protocol in the heat note, and to have you drink to thirst rather than to a schedule.

What runnable chalk lets me see

Runnable ground gives clean watch data, so a Stones build gets good durability reads. For a first 100K the question is simple: does pace at the same heart rate hold past the point where your marathon used to end? Most marathon runners have never run a fourth or fifth hour. When they do, I compare it with the first. I also look at whether a walk break brings heart rate back down. If it stays high, I note it and we talk. Heat, poor sleep, a missed meal or a hard week could all lift it, and the watch can't tell me which.

From marathon to Avebury

For a first 100K I use 24 weeks, the top of my range, so a July race starts in late January. I want a 50K finished before week one. Two strength sessions a week run through it all, never in the 48 hours before a long run.

• Walks from the first hour. Every long run has planned walks on the rises from the start, under the heart-rate cap. The long run grows about 10% at a time against your longest of the previous 30 days, and holds after three steps up.
• The second-morning weekend. From week 15, a back-to-back every two to three weeks, the Sunday on rutted chalk. Sunday starts at 40% of Saturday's time, then 60%, then 70%. The biggest, about 4 hours then 2.5, lands five weeks out.
• The fuelling timer. Start long runs at 30 g of carbohydrate an hour and add 10 g every one to two weeks, so your race rate is settled two weeks out.
• The heat block. Ten to fourteen days of hot baths or sauna after easy runs, ending two or three days before the race.

The taper is three weeks: volume down a quarter, then 40 per cent, then 60, with every run kept. The progression note explains the 30-day rule and the carbs note sets the race rate.

On a Stones build, I watch

• Pace at the same heart rate in hours four and five of long runs, against hour one.
• Whether heart rate comes back down during walk breaks late in a long run.
• The longest run each week against the longest of the previous 30 days.
• Carbohydrate per hour once a run passes three hours, and whether the timer was followed.
• Feet after back-to-back weekends, and whether a planned sock change helps.
• For the two-day format, how the second morning feels against the first.

Facts checked October 2026 against thresholdtrailseries.com (the event and route pages), the event guide PDF that describes the 2025 edition, and Threshold's 2026 charity information sheet: 100 km from Lewknor to Avebury (100.21 km in the 2025 guide), 1,350 m of ascent split 720 m and 630 m by day, the four formats, the 10 to 11 July 2027 dates and fees, red waymarking reflective at night, seven pit stops with water points and the 2025 distances above, wave starts by predicted finish time, the 5 kg and 10 kg bag limits, 40% of the 2025 field hiking or mixing walking and jogging, and a completion rate of over 90%. The organiser's pages give the non-stop limit as both 25+ and 30+ hours, and the 2025 guide's own timings allow longer than either, so check with Threshold before you plan to the minute. A secondary listing gives 1,129 m of ascent; I've used the organiser's 1,350 m. Descent and high point weren't published. The finish and stopping counts are mine, from Threshold's non-stop results for 2023 to 2025 on resultsbase.net, checked October 2026; 2026 results weren't listed. I counted anyone without a finishing position as a non-finisher, and a missed timing mat could put a runner in the wrong place, so treat the stopping points as approximate. The heat, the rutted second half (a blog post of 14 July 2026) and the pacing mistakes come from runner reports, not the organiser. Evidence level: the pacing, fuelling and heat advice matches the field notes. The 24-week build, the back-to-back steps, the fuelling ladder, the heat block timing and the taper cuts are my coaching method, not trial results; the 30-day long-run guardrail rests on one large observational study.