Guide
Marathon Pacing Strategy: How to Run Your Best 42.2 km
The marathon is the only common race distance where pacing isn't just about comfort — it's about chemistry. Run a 10K thirty seconds too fast at the start and you'll fade; run a marathon thirty seconds per kilometre too fast at the start and you can be reduced to a walk with 10 km still to go. This guide explains why the distance punishes optimism so brutally, how to choose a goal you can actually hold, and gives you three complete pacing plans — even, slight negative, and 10-10-10 — with real split tables for a 4:00 goal you can adapt to any target.
Why the marathon punishes optimism: the glycogen math
Your muscles and liver store roughly 400–500 g of glycogen — call it around 2,000 kcal of fast-access carbohydrate fuel. Running costs roughly 1 kcal per kilogram of body weight per kilometre, so a 70 kg runner spends about 2,950 kcal covering 42.2 km. The books don't balance: you cannot run a marathon on stored carbohydrate alone.
What saves you is that you don't run on pure carbohydrate — at easy and moderate intensities your body burns a mix of fat and glycogen, and fat reserves are effectively unlimited. The catch is that the mix depends steeply on intensity. Every notch faster shifts the blend towards glycogen; run just 10–15 seconds per kilometre above the pace your training has prepared you for, and you're draining the small tank disproportionately faster while the effort still feels easy. That's the cruel part: the fuel cost of a too-fast start is invisible for two hours, because the first half of a marathon at almost any sane pace feels comfortable. The bill arrives at 30–32 km, when glycogen runs low, your body is forced to rely on (slower-burning) fat, and pace collapses — the wall. Mid-race fueling raises the ceiling, but it can't refill the tank as fast as hard running empties it. Pacing, not willpower, is what decides whether you meet the wall.
Choosing a goal you can actually hold
The starting point is a recent race result, not a wish. Put your latest half marathon (or 10K) into the race time predictor: it uses the Riegel formula to scale your performance across distances. A 1:52:00 half marathon, for example, predicts a 3:53:31 marathon.
Now the caveat that matters more at this distance than any other: Riegel assumes your endurance is trained in proportion to the target distance. The formula extrapolates from what you did; it can't see whether you've actually run the 30+ km long runs that make marathon endurance real. If your buildup was thin — long runs topping out under 28–30 km, low weekly volume — treat the prediction as a best case and add 5–10 minutes. A 4:00 goal off a 1:52 half is honest for a first or lightly-trained marathon; 3:53 assumes the full buildup went well. Choosing the conservative number costs you three minutes of bragging rights; choosing the optimistic one can cost you twenty at the wall.
The three pacing plans
- Even splits. One pace, start to finish — for a 4:00 goal that's 5:41 /km (use the pace calculator to convert any goal time to its pace). Physiologically the cheapest way to cover the distance, and the plan every world-record attempt is built around. Its weakness is practical: kilometre 38 at 5:41 costs far more effort than kilometre 3, and holding pace on tired legs takes discipline you have to rehearse in training.
- Slight negative split. Start a touch slower than goal pace and finish a touch faster — the table below uses a 2% bias, which for 4:00 means opening at 5:45 /km and closing at 5:38 /km. You bank energy rather than time in the first half and spend it where the race is actually decided. This is how most well-executed amateur marathons look, and it's the shape of Kelvin Kiptum's 2:00:35 world record — more on that in the even vs negative splits guide.
- 10-10-10. The classic mental segmentation: 10 miles with your head, 10 miles with your legs, 10 km with your heart. One concrete way to run it, shown in the table: first 16 km deliberately held back at 5:45 /km, middle 16 km locked on goal pace 5:41, and the last 10.2 km at whatever you have left — on paper 5:35 /km. Its strength is psychological: the race becomes three short races, and the hard part starts exactly where marathons are won or lost.
The same 4:00 goal, three ways
Even and negative-split columns come straight from our splits calculator (negative = the split-bias slider at 2%); the 10-10-10 column uses the three pace bands above. All three finish in exactly 4:00:00.
| Checkpoint | Even (5:41 /km) | Negative 2% (5:45→5:38) | 10-10-10 (5:45/5:41/5:35) |
|---|---|---|---|
| 5 km | 28:26 | 28:41 | 28:46 |
| 10 km | 56:53 | 57:19 | 57:33 |
| 15 km | 1:25:19 | 1:25:52 | 1:26:19 |
| 20 km | 1:53:45 | 1:54:21 | 1:54:49 |
| Halfway | 2:00:00 | 2:00:36 | 2:01:04 |
| 25 km | 2:22:12 | 2:22:47 | 2:23:16 |
| 30 km | 2:50:38 | 2:51:08 | 2:51:42 |
| 35 km | 3:19:05 | 3:19:25 | 3:19:50 |
| 40 km | 3:47:31 | 3:47:38 | 3:47:45 |
| Finish | 4:00:00 | 4:00:00 | 4:00:00 |
Notice how small the differences are on paper: at halfway, the most conservative plan trails the even plan by just 64 seconds. That's the whole point — a good marathon pacing plan looks almost boring in a table, because the real difference happens in your fuel tank, not on the clock. To generate the same table for your goal time, enter the distance, target time and a bias into the splits calculator and it recomputes every kilometre instantly.
Fueling checkpoints, tied to your pace band
Whatever plan you run, fueling is part of pacing — a gel you skip at 90 minutes is pace you lose at three hours. A practical rule is carbohydrate every 35–40 minutes, starting early; at 4:00-goal pace, every 40 minutes is roughly every 7 km:
- ~7 km (0:40): first gel — yes, this early. You're fueling kilometre 30, not kilometre 7.
- ~14 km (1:20): second gel; check in on pace — are you within a few seconds of your band?
- ~21 km (2:00): third gel at halfway; compare your actual split to the table's.
- ~28 km (2:40): fourth gel, before the hard part starts.
- ~35 km (3:20): last gel if your stomach allows; from here it's pace discipline.
Drink to thirst at stations rather than to a schedule, and practise the exact gels and spacing on your long runs — race day is not the place to discover your stomach disagrees.
When the plan breaks at 30 km
Sometimes it breaks anyway — heat, a stomach that stops cooperating, a course that's hillier than the profile suggested. What matters is what you do in the sixty seconds after you realise it:
- Downshift once, deliberately. Ease off by 15–20 seconds per kilometre — a controlled step down to a pace you can hold, not a slide. One planned downshift usually saves the race; three unplanned ones is the collapse you were trying to avoid.
- Fuel and drink immediately. Early-wall symptoms are often a fuel problem wearing a fitness costume. A gel plus water can bring a dying pace back to life within ten minutes.
- Re-anchor to the new target. 4:00 becomes 4:06 — take ten seconds to accept it and recompute your remaining splits at the new pace (roughly +90 seconds per remaining 5 km block for a 20 s/km downshift). A recalculated goal keeps you racing; a lost one turns the last 10 km into survival.
- Never try to win it back at once. If the crisis passes and the legs return, come back to goal pace — but don't try to repay a three-minute debt in two kilometres. That's how one wall becomes two.
Common mistakes
- Banking time in the first half. "I'll get ahead of schedule while I feel fresh" is the single most reliable way to blow up a marathon. Feeling fresh at 15 km is not information — everyone feels fresh at 15 km. The glycogen ledger doesn't care how it felt.
- Starting with the wrong crowd. Lining up ahead of your pace group and getting swept along at someone else's rhythm can cost you 10 s/km without you noticing. Start behind your goal-pace group and let the first kilometre be your slowest.
- Racing the GPS instead of the course. GPS pace wobbles in cities and tunnels, and every weave around other runners adds distance — most runners' watches show 42.5+ km by the finish. Check your watch's cumulative time against the course's marked kilometres — that's what the split table is for — rather than reacting to instant pace.
- A goal with no evidence behind it. Round numbers are seductive: 3:59:59 is not a training outcome, it's a wish, unless a recent race run through the race time predictor says it's in range. Pick the goal from data, then pick the plan.
- Trying a new plan on race day. Whichever shape you choose, rehearse it — the middle third of your longest runs at goal pace, fueling on schedule, even the downshift protocol. The plan you've practised is the only one you'll execute under fatigue.
Build your own split table
Set your distance, goal time and split bias in the splits calculator to get your personal version of the table above, kilometre by kilometre. Convert goal times to paces with the pace calculator, and sanity-check the goal itself against a recent result with the race time predictor. Then write the checkpoints on your arm, and run the boring plan — boring tables make happy finish lines.