The most common question in marathon training: how many kilometres per week should I be running?
Most training plans give you a number. But where do those numbers actually come from? We dug into published research on real runners' training data — including analysis of thousands of amateur marathoners' actual training logs — to find out what the data shows about volume and performance.
The Volume-Performance Relationship
The relationship between training volume and marathon finish time is strong — but it plateaus faster than most runners expect.
Research analysing the training patterns of recreational marathon finishers consistently shows a clear pattern across finish time brackets:
| Goal Finish Time | Recommended Peak Week | Minimum Effective Volume |
|---|---|---|
| Sub 3:00 | 80–100 km | 70 km |
| Sub 3:30 | 60–80 km | 50 km |
| Sub 4:00 | 50–65 km | 40 km |
| Sub 4:30 | 40–55 km | 30 km |
| Sub 5:00 | 30–45 km | 25 km |
| Finish Only | 25–35 km | 20 km |
The "minimum effective volume" column is important: it represents the weekly mileage below which the data shows a significant increase in injury rate, underperformance, and hitting the wall.
Key Takeaway
For most recreational runners targeting sub-4:30, a peak week of 45–55 km is the sweet spot. Going above 65 km without years of adaptation delivers diminishing returns and sharply increases injury risk. Going below 30 km makes the final 10 kilometres very hard.
The Diminishing Returns Curve
The data shows a non-linear relationship between weekly volume and performance. Going from 30 km/week to 50 km/week delivers much larger improvements than going from 60 km/week to 80 km/week.
This has practical implications:
- The biggest gains come early. If you're currently running 25 km/week, adding 10 km/week will likely improve your marathon time by 15–25 minutes.
- Elite volumes don't apply to recreational runners. Professional marathoners run 150–200 km/week, but the data on recreational runners shows minimal additional benefit beyond ~80 km/week for non-elite athletes.
- Consistency beats peaks. Three months at 45 km/week consistently outperforms a plan that spikes to 65 km/week but includes injury breaks and forced rest weeks.
Research suggests that time on feet — how many minutes you run per week — is more predictive of marathon performance than raw kilometre count. Running slowly at high volume is not the same training stimulus as running moderately at moderate volume.
The Long Run: How Long Is Long Enough?
The long run is the cornerstone of marathon training, but there's persistent debate about how far it should go. The data on recreational runners points to a clear answer:
For most runners targeting 4:00–5:00 finish times, the long run sweet spot is 28–34 km (17–21 miles), run at conversational pace.
Going beyond 34 km in training runs has not been shown to improve marathon performance for recreational athletes. It does significantly increase recovery time and injury risk.
| Finish Time Target | Recommended Long Run Range | Max Recommended |
|---|---|---|
| Sub 3:30 | 30–35 km | 37 km |
| Sub 4:00 | 28–34 km | 35 km |
| Sub 4:30 | 26–32 km | 33 km |
| Sub 5:00+ | 24–30 km | 32 km |
How to Structure Your Weekly Volume
Total weekly kilometres matter less than how they're distributed. Research on amateur runners consistently shows that the most effective weekly structure follows roughly:
- 80% of weekly volume at easy/conversational pace — this builds aerobic base, develops fat oxidation, and recovers the body
- 15% at moderate/tempo effort — threshold runs and marathon-pace work
- 5% at hard effort — intervals, track work, or race pace segments
A runner doing 50 km/week with this distribution will almost always outperform a runner doing 55 km/week at uniform moderate pace.
The most common training mistake in the data: running every session at moderate (comfortable-hard) effort. This sits in a "junk miles" zone — too easy to stimulate maximal aerobic development, too hard to allow recovery. True easy running should feel embarrassingly slow.
The Taper Effect
The last 2–3 weeks before a marathon, the data consistently shows runners should reduce volume by 30–40%. This is called the taper, and the research behind it is robust.
Taper-phobia — fear of losing fitness by reducing mileage — is common but unfounded in the data. Fitness does not decline measurably in a 2–3 week taper. What does happen is glycogen restoration, muscle repair, and neural recovery that allows race-day performance to peak.
Most training plans implement too short a taper (1 week) or not enough volume reduction. The data supports a 3-week taper at approximately this pattern:
| Week Before Race | Volume Reduction |
|---|---|
| Week 3 out | 30% reduction |
| Week 2 out | 50% reduction |
| Race week | 70% reduction (with short sharp effort mid-week) |
Data sources: Published sports science research on recreational marathon training volume; Strava Global Running Report; academic literature on marathon preparation and performance outcomes.