Goal time · Marathon
A 3:30:00 marathon works out to 8:01 per mile, or 4:59 per kilometer. Below are the mile splits for a slightly faster second half, the halfway time to aim for, and the training paces that go with this goal.
The numbers
This plan starts about 1% slower than goal pace and finishes about 1% faster, so the second half is slightly quicker than the first. The elapsed column is what your watch should read as you pass each marker.
| Mile | Pace for That Mile | Elapsed Time |
|---|---|---|
| 1 | 8:06 | 8:06 |
| 2 | 8:06 | 16:12 |
| 3 | 8:05 | 24:17 |
| 4 | 8:05 | 32:22 |
| 5 | 8:04 | 40:26 |
| 6 | 8:04 | 48:30 |
| 7 | 8:03 | 56:34 |
| 8 | 8:03 | 1:04:37 |
| 9 | 8:03 | 1:12:39 |
| 10 | 8:02 | 1:20:41 |
| 11 | 8:02 | 1:28:43 |
| 12 | 8:01 | 1:36:44 |
| 13 | 8:01 | 1:44:45 |
| Halfway (13.1) | 8:01 | 1:45:38 |
| 14 | 8:00 | 1:52:46 |
| 15 | 8:00 | 2:00:46 |
| 16 | 8:00 | 2:08:45 |
| 17 | 7:59 | 2:16:44 |
| 18 | 7:59 | 2:24:43 |
| 19 | 7:58 | 2:32:41 |
| 20 | 7:58 | 2:40:39 |
| 21 | 7:57 | 2:48:36 |
| 22 | 7:57 | 2:56:33 |
| 23 | 7:56 | 3:04:29 |
| 24 | 7:56 | 3:12:25 |
| 25 | 7:56 | 3:20:21 |
| 26 | 7:55 | 3:28:16 |
| Finish (26.2) | 7:55 | 3:30:00 |
Measured against the latest published results of 214 marathons on RunSesh.
One edition per race, from RunSignup races with published results. Fields include walkers and charity entrants, so this describes these races, not every runner in the country.
These are the paces for a runner who can already race a 3:30:00 marathon, worked out with Daniels and Gilbert's VDOT formulas, the method behind the training tables in Daniels' Running Formula. If your most recent race is slower, train at the paces on the page for that time.
Most of your weekly miles, including the long run.
A steady run of about 20 minutes, or longer repeats with short rests.
Repeats of 3 to 5 minutes, with easy jogs between.
Smyth's 2018 analysis of 1.7 million recreational marathoners, published in the Journal of Sports Analytics, found that starting too quickly can result in poorer finish times, which shapes why this plan builds a slightly faster second half. Treat the halfway marker as your anchor on race day: if you arrive there feeling controlled, the back half is where the plan pays off.
Daniels and Gilbert's VDOT formulas treat a 45:37 10K as the same performance as a 3:30 marathon, which gives you a benchmark you can test well before race day.
The RRCA pace chart is pointed on this: "don't begin with a future goal time. To do so increases risk of injury and/or overtraining." Base your paces on a race you've already finished.
USATF certification measures courses along the "Shortest Possible Route," so the certified distance is a measured minimum. Pobiruchin et al. (2017) found GPS sport watches averaged 42.33 km against the official 42.195 km, with watches recording 135 meters extra, worth 40 seconds at this goal's pace. If your watch logs that much additional distance, it could show a faster average than 8:01 even when your effort is exactly on goal.
Daniels and Gilbert's formulas treat a 1:41:06 half marathon as equivalent to a 3:30 marathon. Arriving at that time in a recent half gives you a grounded way to gauge whether your aerobic fitness currently supports the full-distance goal, rather than relying on a projection.
Ely et al. (2007) studied seven marathons and found performance slowed progressively as WBGT, a heat-stress index combining temperature, humidity, sun, and wind, climbed between 5 and 25 degrees Celsius, with runners finishing between 25th and 300th place losing more time than faster runners. El Helou et al. (2012) analyzed 1,791,972 finishing performances and found that above each performance level's optimum air temperature, speeds fell and withdrawals rose.
