If you are chasing a sub-4 marathon — roughly 5:41 per kilometre, or 9:09 per mile — the weekend long run is the session that does most of the work. It is where your body learns to handle 42.2 km without falling apart, and it is also, as it turns out, the session where most of the injury risk lives.
Structuring the long run
Most sub-4 plans build the long run up to 32–35 km (20–22 miles) over 16 to 20 weeks. This is standard coaching practice rather than a research finding, but it is standard for a reason: it gets you within touching distance of race distance without racing it every weekend. The key is running it slow enough to stay aerobic.
- Pace: 10–20% slower than goal marathon pace. For sub-4, that is roughly 6:20–6:40 per km (10:10–10:45 per mile). It should feel conversational.
- Heart rate: mostly in Zone 2, around 60–70% of maximum, if you train by heart rate.
- Finish strong: once every two or three weeks, run the last 5–8 km at goal marathon pace to rehearse the feel of racing on tired legs.
Progression — and the number that actually predicts injury
This is the part worth getting right, because it is the part with hard evidence behind it.
A 2025 study in the British Journal of Sports Medicine tracked 5,205 runners across 588,071 GPS-recorded sessions. It found that the strongest predictor of overuse injury was not weekly mileage, and not the acute-to-chronic workload ratio. It was a single run that exceeded the runner's longest run in the previous 30 days by more than 10%. Runs 10–30% longer than that benchmark carried a 64% higher injury rate; runs more than double it, a 128% higher rate.1
For the weekend long run, that translates directly:
- Increase the long run by no more than 10% of your longest in the last 30 days. At 25 km that means adding at most 2.5 km; at 32 km, at most 3.2 km. The old coaching rule of "2–3 km every 2–3 weeks" sits comfortably inside this, which is probably why it survived.
- Every third or fourth week, drop the long run back by 20–30%. This is convention rather than something the study tested, but it also resets your 30-day benchmark to something you have already handled.
- The temptation to "add a bit" on a good day is exactly the spike the study measured. Feeling strong is not a signal to extend; it is a signal that the plan is working.
We have written up the study in more detail, including why it changed the old 10% rule: How much running is too much?
Fuelling the long run
Practise race nutrition on every long run over 90 minutes: gels or drink every 30–45 minutes, starting around the 45-minute mark. The point is to train your gut as well as your legs, so that nothing on race day is a first attempt.
One more lever
Load management is the first line of injury prevention for the long run, and the evidence for it is as good as running research gets. The second lever with comparable evidence is foot strengthening — a year-long randomized trial found runners who trained their feet were 2.4 times less likely to get injured. It is a separate intervention and it does not buy you headroom above the 10% line; if you want the detail and the limits, it is on the runners' page.
References
- Schuster Brandt Frandsen J, Hulme A, Parner ET, Møller M, Lindman I, Abrahamson J, Sjørup Simonsen N, Sandell Jacobsen J, Ramskov D, Skejø S, Malisoux L, Bertelsen ML, Nielsen RØ. How much running is too much? Identifying high-risk running sessions in a 5200-person cohort study. British Journal of Sports Medicine. 2025;59(17):1203–1210. doi:10.1136/bjsports-2024-109380. Read it
