How to Predict Your Marathon Time

Let’s say you’re cruising through your first marathon.  You’re running a little faster than you’ve trained, but hey, it’s race day, and you’re charged with adrenaline.

You cross the halfway point five minutes faster than you had planned based on your training pace.  You’re rocking your first race, and everything is great.  And then disaster strikes.

At Mile 15, it hits you.  For the first time, you notice your legs feel heavy.  By Mile 16, the idea of a quick walk break sounds great.  By 17, you’re not sure you can even finish the race.

What happened?  You were so disciplined in your training, and you tried to listen to your body on race day.  The problem is that you didn’t know what you were capable of and had no idea how fast to run.  And a good rule of thumb is that for every minute too fast that you run the first half of a marathon, you’ll lose two minutes in the second half.

You need a better idea of how fast you can reasonably expect to run.  These three marathon prediction methods well help you do that, and avoid the race-day disaster that so many runners experience.

Yasso 800’s

Here’s the idea.  Pick a marathon time you’d like to run.  Replace hours by minutes and minutes by seconds to get your 800 meter pace.

Sounds complicated, but it’s not!  Since I need to run the marathon in 3 hours, 10 minutes to qualify for Boston, my 800 pace becomes 3 minutes, 10 seconds.  If you can do 10 of these 800’s, slowly jogging for that same amount of time between each (3:10 for me), then this method predicts you can run your marathon at the corresponding hours-and-minutes time!  A lot of people say that Yasso 800’s are too optimistic a prediction time, so adding five minutes to the time you predict is probably a good idea.

Here’s a link to Runner’s World’s take on Yasso’s.

Galloway’s Magic Mile

Former Olympian and current coach Jeff Galloway (the walk-run guy) has his own method for predicting race times, and not just marathons.  And the great thing about his method is that it’s simple, albeit painful.

Here’s how it works.  Run one mile as fast as you can, middle-school style (don’t barf or die, though).  Then to predict your marathon time, take your mile time (in seconds) and multiply by 1.3.  The result is your marathon pace per mile, in seconds.

I haven’t done this one yet, but just to give an example:  In my last 5K, I ran the first mile in 5:53.  Let’s say that if I were going all-out, I could have run a single mile in 5:40 (you’re supposed to actually do it, of course, not guess).  That’s 340 seconds, and 340 x 1.3 = 442 seconds, or 7:22 marathon pace, for a 3:13 marathon.

Galloway actually gives you a break; he says to run four one-mile time trials, drop your lowest, and average the remaining three to get your mile time.  He also gives different adjustment factors to predict 5K, 10K, and half marathon times at his website.

[daniels running formula image]Jack Daniels’ Tables

Jack Daniels is a well-known running coach and author of a book I really like, Daniels’ Running Formula (Amazon affiliate link).  In the book, he gives a table for predicting any common race time, given any other common race time, based on what seems like a lot of research and statistics.  You can find such a pace table here (Table 1).

Basically, you look up a recent race time, and from that you look to the left or right to find your projected time for another distance.  Don’t worry about the “VDOT” column, it’s something from the book.  Ideally, you want the race distances to be similar (i.e. use a half marathon time to predict a marathon time, if you have a recent one).  You also want things like the terrain and conditions to be similar.

Example:  My half marathon a few months ago was extremely hilly, so I can’t really compare that one.  But if I look up my recent 5K time of 19:15 on a very flat course, the equivalent marathon time is about 3:04.

If you’re not into tables, running coach Greg McMillan has a similar tool here, called McMillan’s Running Calculator.

This post is part of a series of posts designed to teach you how to run long and strong.  Go check out the rest!



Dig this post?
Spread the word!

Keep in touch:

The 7 Foods Worth Eating Every Single Day

wooden signpost near a pathOur 7-Day Kickstart Plan is unique in that it focuses on the highest quality whole foods (including the 7 foods worth eating every day), to make sure you get everything you need on a plant-based diet.

The Kickstart Plan includes:
  • A 7-day meal plan, built around the foods worth eating every single day
  • 14 of our favorite recipes that pack in the nutrition, taste great, and are easy to make
  • Focused on simplicity and speed, to minimize stress and time commitment
It's the best way we know of to get started with a whole-food, plant-based diet, for just 7 bucks. Learn more here!


  1. wow a 3:07! get it!! 🙂

    i am always tinkering with the online calculators. i did yasso 800s last fall/winter in prep for the ’09 goofy challenge. i was running them 3:20 or better. i am not sure how effective they are as a measurement though because i was doing the half and the full. i went all out saturday in the half and just tried to survive the full (successfully). i ran a 1:34 half though, i wonder if that can be ‘linked’ to my sub-3:20 800s somehow?
    .-= lindsay´s last blog ..we got ourselves a game =-.

  2. Congrats on the Goofy Challenge. Crazy if you ask me, though! My wife and I did the Disney full this year, maybe we saw you!

    A very rough estimate is to double the half and add 5 minutes. So by that, your 1:34 half is “equivalent” to a 3:13 full. So sub 3:20 Yasso’s makes sense, I guess!

  3. Okay… so with a marathon coming up soon, I can see why you’d want to keep at it for consistent training… but you don’t want to overdo it and hurt yourself! Maybe taking a couple of rest days would actually be beneficial to your training?
    .-= Sagan´s last blog ..Product Review: Glee Gum =-.

    • NoMeatAthlete says:

      Hey Sagan, thanks for being concerned 🙂 If this were a legitimate injury, I’d definitely consider more rest. But it’s just a blister, so as long as it’s not causing me to change my form, I don’t think there’s risk of getting hurt, like there would be running with an injury. Unless the blister never heals!

  4. I haven’t done Yasso 800s before, but I have “reverse-engineered” all of the other methods. In that, when I heard about them, I went through my training logs, looked for results in races or workouts for the various distances, plugged them in to the formulas, got my marathon time, and then compared against the actual marathons.

    They are usually relatively close.
    .-= Blaine Moore´s last blog ..Emily’s Run 5k (2009) =-.

    • That’s good to know. I’m a little doubtful about the Yasso 800’s. For example, yesterday’s workout (part of my program to run a 3:10 marathon) was supposed to be 6 of them at 2:52, with only minute and a half rests. I would think that if you could complete this workout, you could do much better than 10 800’s at 3:10 with 3:10 rests. So that would mean either the program or Yasso got something wrong.

  5. Based on my 5K time of 19:41, Daniels predicted a half-marathon of 1:30:22, and I ran 1:30:19. That demonstrated to me that Daniels is onto something and that I was well trained to run the longer distance. That latter is key to these things being meaningfull, of course. (If I put all my time and energy into running a 18:41 5K, and succeeded, I doubt I would then find myself shaving 20 minutes off my marathon PR and hitting the Daniels-predicted 2:58:52. Heck, I probably wouldn’t make four hours…)
    .-= Pete´s last blog ..Getting to Where I Want to Go =-.

  6. Plugging my 15K and 5K time into the Daniels tables I would hit my 3:40 that I am looking for. Do I think it is going to happen, I am not sure. In the past I have used the McMillan calculator, but I have read that it is more based on higher 70+ miles, in order to make it more accurate. For my shorter races, I seem to be fairly consistent for the times it predicts. I just need to get my endurance up there in order for my marathon to be consistent with the other race times.

    My husband and I also did the Disney Marathon this past year. Our second year in a row.

    And I got my shirt today…I will be wearing it on my run tomorrow and will take a picture and post it after.
    .-= Robin´s last blog ..Posts are so much more fun with pictures =-.

    • Robin, I’m confused about the “70+ miles” thing. How is the calculator based on that? People who run 70+ per week?

      I can’t wait to see your shirt picture!

      • Hi
        Most of the online race calculator are based on the Riegel formula, which adds a constant slowdown, every time the distance doubles. This exhaustion factor works well for races up to the half marathon or so. But it does not take into account depleting glycogen levels, which of course is important on a marathon (unless you are in real great shape;).
        Especially when running the first marathon, I would not recommend going out faster than training long runs, no matter what these tools say.
        Cheers, Peter

  7. I used Galloway’s Magic Mile method for training and to predict my half-marathon pace. A very effective way of setting your pace.
    .-= kara´s last blog ..Where Do You Pin Your Bib? =-.

  8. Christof Schwiening says:

    Obviously this is a long dead thread – but, Tanda (2011) provides a very good estimate of marathon performance from training data. It is a simple formula involving just pace and distance covered over 8 weeks starting from 9 weeks before the marathon. Whilst Tanda’s original dataset is quite small I have tried the formula on a range of both my runs and those of others. It predicts my times almost perfectly (within a couple of mins) and for about a dozen other runners I have found on Strava with finishing times from 2:40 down to 3:45. Of course if training is done on hills, hot weather or heavy shoes then it will be an underestimate of performance – equally it is possible to fail to achieve the Tanda prediction by going off too fast or running a hillier marathon than the training profile. Nevertheless it is by far the best predictor I have come across for someone trying a new training program.

  9. In my experience Yasso 800s are too optimistic and Galloway’s Magic Mile is too pessimistic while Daniel’s prediction is pretty close.

    Last spring I completed a session of ten 800 meter repeats at an average of 3:00 but finished my marathon in 3:11. My mile time of 5:50, however, would have predicted a finish time of 3:18. Using recent race results of various distances and using the VDOT value from Jack Daniel’s table gave the best predictions.

  10. For me, this thread is far from dead! Thank you for some much-needed advice. Today I failed to reach my goal distance (34K) just eight weeks out from my first marathon: heavy legs at 25K, massive and unexpected doubts at 27K, exactly as you described. I don’t often give up but I think I made the right decision to quit at 30K, not least because I am 60 years old. So I’m well-chastened for not respecting the distance. Also, it proved surprisingly difficult to resume running after a compulsory but less-than-ideally timed break for a national anthem…Today’s lesson is a valuable one, and I wonder what new mistakes I will make next time…and in my 30 – 41K blind-spot. Now, if I can only find running shoes that fit my big feet (not so easy in SE Asia)….BTW, I didn’t know there were pacers in these events.

  11. so if i want to run 3:45 at the MCM, i need to do ten 800s at 3:45, sounds too easy if you ask me.

Leave a Comment