A single recent performance is enough to estimate what you're worth over the other distances, and to derive the paces you should be training at. The calculator below does it live.
One warning first, because it changes how to read the result: these numbers are a statistical estimate, not a promise. They say what a runner with your speed and a coherent build-up usually achieves. Not what you'll achieve on Sunday.
Enter a recent performance to see the projections.
What it projects over the other distances
| Distance | Projected time | Pace /km | Speed |
|---|
The training paces that follow
| Zone | Pace /km |
|---|
What the calculator computes, and how
It applies Riegel's formula, which links two performances through a power law: the time over the new distance equals the old one multiplied by the ratio of distances, raised to the power of 1.06.
That small exponent carries the whole meaning. Doubling the distance doesn't double the time, it multiplies it by slightly more than two, because pace degrades a little as the effort lengthens. In practice, 25 minutes over 5K gives about 52 minutes over 10K, not 50. A 45-minute 10K runner projects around 1:40 for a half.
It's an empirical law, fitted on tens of thousands of race results. It models no physiology: it observes a regularity. That's its strength, because it works for almost everyone, and its limit, because it knows nothing about you in particular.
Where the projection gets it wrong, and by how much
The formula assumes your endurance matches your speed. That assumption holds from one distance to the next. It collapses beyond that.
A runner who only races 5K sees their projected marathon come out ten to twenty minutes optimistic, sometimes more. The reason is simple: nothing in a 5K says whether you'll hold up for three hours. Speed gets measured over a short distance. Endurance cannot be inferred from it.
The practical rule fits on one line: trust one step up, treat two steps as an ambition. A 10K projected from a 5K is reliable. A marathon projected from a 5K is an intention.
And the projection ignores everything else: heat, elevation, wind, the week's sleep, whether the fuelling worked. A marathon run at 28 °C, 82 °F, costs several minutes whatever the formula says.
The training paces that follow
The second table doesn't come out of nowhere: each zone is derived from a projection, with a stated rule.
Easy running sits at 10K pace plus 60 to 75 seconds per kilometre. That's where most of the volume lives, and it's the zone almost everyone runs too fast. If conversation doesn't flow in full sentences, it's too quick.
The long run holds 10K pace plus 45 to 60 seconds. Slightly firmer than an easy run, never hard.
Marathon pace is the marathon projection itself. Threshold matches the half projection, which is roughly the pace you can hold for an hour. Intervals of three to five minutes get run at projected 3K pace.
One last point, and it outranks the other four: these paces are ranges, and sensations beat the number. At 30 °C, 86 °F, on a hilly route or after a bad night, the right pace for the day is slower than the table says, and following it anyway is the most reliable way to turn a useful session into a lost week.
Reference points, if you have no recent performance
Without a timed race, this table gives an entry point. The bands are deliberately wide.
| Profile | 5K | 10K |
|---|---|---|
| First race completed | 32 to 36 min | 1:08 to 1:18 |
| A year of regular running | 25 to 28 min | 52 to 58 min |
| Three or four runs a week for years | 21 to 23 min | 44 to 48 min |
| Strong club level | 18 to 19 min | 37 to 40 min |
What the number is actually for
Two uses, and one trap.
The first use is choosing a race target that stands up, then starting at that pace instead of starting on the sensations of kilometre one, which always lie because everyone is fresh and the crowd pushes.
The second is dating your progress. The same reference performance re-measured three months later tells you whether the training is paying, without needing another race.
The trap is racing at the projected pace when the training was done at another. A projection is not a preparation. A 1:45 half projected from a good 10K stays out of reach if the longest run of the quarter was 12 kilometres, a bit over 7 miles.
Two numbers to keep in your head alongside it
The calculator assumes something it never says out loud: that you arrive rested. Every projection describes a runner who tapered, slept and turned up fresh. Feed it a performance set in the middle of a heavy training block and the output is conservative; race off the back of a heavy block and the output is fiction. The projection is a ceiling for a good day, not a floor for any day.
The second number is the gap between projected pace and goal pace. On a distance you've never raced, subtracting five seconds per kilometre from the projection and holding that for the first half is what produces a good finish. Runners who start exactly on the projected pace tend to arrive at three quarters of the distance with nothing left, which costs far more than the five seconds ever would.
FAQ
Is Riegel's formula reliable?
Over neighbouring distances, yes, within a few percent for the large majority of runners. It degrades as the ratio between the two distances grows, and it turns frankly optimistic towards the marathon. It's a solid estimating tool and a poor oracle.
Why does my projected marathon look too fast?
Because the formula reads your speed and ignores your endurance. It assumes a build-up consistent with the reference performance. Someone whose weekly volume tops out at 30 kilometres, under 20 miles, will see a marathon projection well beyond what their legs know how to hold over the full distance.
Which performance should I use as the reference?
The most recent one, over the distance closest to your goal, run in a race or solo at full effort. A 10K race beats a 5K training run, and anything older than three months counts mostly as history.
Should I correct for elevation or heat?
Yes, and by hand. No projection formula accounts for terrain or temperature. A hilly course and a hot day each cost several percent, and together they stack. The useful reflex is to aim at effort instead of pace whenever conditions are out of the ordinary.
Can I project from a training run instead of a race?
Yes, knowing the number will be pessimistic. Nobody digs as deep alone on a Tuesday evening as they do on a start line. The gap runs around two to four percent, more for runners who need the atmosphere to go past themselves.
Should my training paces follow the table or my sensations?
Sensations, with the table as a guardrail. Its real use is catching gross errors: an easy run done fifteen seconds per kilometre too fast, week after week, is what plateaus the most regular runners.
Does the 1.06 exponent apply to everyone?
It's an average. Very endurance-oriented runners are better described by a lower exponent, fast and short runners by a higher one. The difference stays marginal between neighbouring distances and becomes visible across a large gap.
Does the calculator work in miles?
Yes, fully. The Metric or Miles switch at the top of the calculator flips paces to per mile and speed to mph. The distance list stays the same, deliberately: a 5K is still a 5K in the States, it's the name of the race. The choice is remembered between visits, and it carries across the other calculators on the site.
A projected number is not a built week
The projection says where you can go. It doesn't say how to get there, and that's where most build-ups come apart: the right target, the wrong week. The Vitality Archetype test measures your profile across 5 dimensions in 27 questions and 3 minutes, then generates a training week calibrated for your recovery, your available time and your relationship with effort. The paces above then become sessions, placed on the right days.