“No long run should last longer than three hours.” This guideline is popular and prolific on social media. For the 3-3.5 hour marathoner, that’s just common sense. But what about the five-hour marathon finisher? Or the ultrarunner who uses a couple 50K races are training runs? What happens if you take 3:20 hours on your 20-mile long run instead of 3 hours? Is the three hour long run rule actually based in science?
This article delves into the evidence for and against the three hour long run rule. We’ll also look at nuances in training to consider when deciding how long to run before your marathon.
You can also listen to the Tread Lightly Podcast – the first episode addressed the topic of the three hour long run rule.
The Three-Hour Long Run Rule
The three-hour rule emerged over the past few decades based on expert opinion. Dr. Jack Daniels argued on the basis of elite practice. Since elite runners do not train over 2.5 hours in a single long run, then recreational runners should not either.
The rationale is that after 2.5-3 hours, metabolic adaptations reach a point of diminishing returns. If an athlete is at a point of diminishing returns and not spending more time on their feet in racing, why extend the long run?
Exercise physiologist and running coach Luke Humphrey has further popularized the three-hour rule. His Hansons Marathon Method peaks at a 16-mile long run in marathon training. Humphrey provides a thorough overview of the expert opinion of Dr. Vigil, Dr. Daniels, and Dr. Noakes. (One caveat: Dr. Noakes has since fallen out of favor in the exercise science and medical community for his extreme stances.)
Research vs Expert Opinion in Exercise Science
You often read “research says long runs should not exceed three hours.” However, no study conclusively states that.
That approach comes from expert opinions (as explained above), which can be valid. However, expert opinions are seldom definitive or infallible. In long-distance running, expert opinions often come with a big asterisk: they come from coaches who work with athletes who are probably finishing marathons in under 3 hours.
Expert opinions are the lowest level of evidence in the hierarchy of evidence used to make evidence-based claims. You need to use peer-reviewed research – and properly interpret and apply that research. You cannot just say “research claims” without citing a source. Exercise science graduate school is where you lose points if you so much as omit a p-value when citing a study.
Instead, evidence-based approaches to training must consider what the research says before proclaiming any training practice as Gospel. And even when the research supports a theory, it supports – we really can’t “prove” these things. We can understand the statistical significance of hypotheses, refute or accept null hypotheses, and create theories – but very few things in physiology are definitively proven.
Research includes randomized controlled trials, meta-analyses, systematic reviews, and longitudinal studies. Each study used must be assessed for the quality of evidence it provides. Studies can support expert opinions, and expert opinions can guide the direction of studies. But to make evidence-based claims in coaching, you need to look at actual studies.
A Quick Preface to the Research on Longest Long Run
So, let’s delve into the research. Keep in mind: there is no study out there directly on a three hour long run. (If there is, someone please share in the comments below.) Additionally, exercise science is a messy field – few things are absolutely definitive. If someone claims the three hour rule is 100% proven based on research, they may be misunderstanding how research works.
One study commonly cited for no additional aerobic benefits beyond two hours of running was published in 1982 – with rats as the test subjects. Rats are used in scientific research to observe biological responses. Dudley et al. (1982) concluded that rats who exercised for 90 minutes per day saw plateauing responses in mitochondrial adaptations compared to those who exercise for 60 minutes – and the plateau was extrapolated to 2 hours and 3 hours.
Most recent studies related to marathon long runs assess training factors (including long run length) in relation to performance. Thus, there may be some limitations. Correlation does not always indicate causation. So, a correlation of a certain long run with a certain finish time does not always indicate that there is statistically significant causation.
What Does the Research Say on Three Hour Long Runs?
A 2020 study in the Scandinavian Journal of Medicine and Science in Sports examines the relationship between the longest endurance run, marathon performance, and injury risk. The researchers found that completing <25km (15.5 miles) as the longest run in marathon training correlated with slower finish times (p<0.01). Lower weekly mileage also correlated with slower finish times. It’s worth noting the study excluded any participants who did not finish, so we don’t know if shorter long runs had an association with not finishing the marathon or if those marathoners completed runs so long that they were overtrained by race day.
Additionally, the whole cause and effect mechanism is not fully clear. Are the runners slower because their long runs were not long enough? Or did they not complete a longer run long because of their slower training pace?
Interestingly, the researchers found no correlation between longest long run and injury risk. The odds ratio of the longest long run of <25 km and injury was 1.00 (OR of 1 means that there is no causal relationship found). A long run of >35 km had an OR of 1.01 with injury (slight correlation), while an long run of 25-30 km had an OR of 0.75 (not associated with injury risk).
A 2020 meta-analysis of five decades of research published in Journal of Science and Medicine and Sport looked at the relationship between training behaviors and marathon performance. The number of 32km (19.9 miles) runs and the longest long run of training were both “significantly associated” with a faster marathon finish time (p<0.001). Peak weekly training volume was also significantly associated with faster finish times (p<0.001). However, based on the r-squared value, long runs of 32km had the greatest predictive capacity on finish time.
The researchers concluded that four-hour marathons do not need to run more than 32km in preparation for the marathon. While not definitive in limiting runs based on time, it does reinforce the notion that adequate preparation -but not excessive prep – is key in the marathon.
The Relationship between Fueling and Muscle Damage
Coaches often limit long runs to three hours based on perceived injury risk. This is a valid concern. Your weekly long run should not beat you up so much that you are unable to complete the following week of training.
Let’s take a critical look at the era in which the 2.5-3 hour long run rule emerged. Even in the most recent edition of Daniels Running Formula, he advises withholding any form of carbohydrates on long runs and just taking in water “occasionally.” No carbohydrate intake over more than 2.5 hours will naturally increase muscle damage, as glycogen depletion will occur around 2 hours.
In the Hansons Marathon Method, Humphrey explains that glycogen depletion is a significant factor in the rationale for limiting long runs to 16 miles. His rational concern is that the rate of glycogen depletion in long runs will diminish the quality of training for the 72 hours following a long run.
Today, we understand more about the factors influencing muscle damage. Muscle damage may be partially due to protein breakdown in the working muscles for gluconeogenesis. If you do not have enough carbohydrates to support the session demands, your body takes protein from your muscles and converts it into glucose for ATP production. Usually, gluconeogenesis contributes to only a small portion of oxidative energy production. However, when substrate availability is inadequate, you have to get that energy from somewhere.
Sport nutrition knowledge has expanded significantly over the past 5-10 years. Now, research such as a 2020 study in Nutrients suggests that high carbohydrate intake (up to 120 g CHO/hr) during long-distance running may reduce exercise-induced muscle damage, as observed in biomarkers such as creatine kinase.
Are the injury risks the same if a long run is supported by a high carbohydrate intake? This is an area for further research. Based on the current evidence though, I hypothesize that higher carbohydrate intake reduces the risk of injury on long runs.
Long Runs, Recovery, and Injury Risk
Boulousa et al. (2020) described the higher training volume as a “trigger” for injury if the runner had increased risk for other reasons. The volume is not an inherent risk, but if someone is already at high risk, it can serve as a tipping point. Instead of making blanket statements that long runs over three hours always increase injury risk, it is worth looking at long runs in the context of an athlete’s overall injury risk:
Individual factors also affect recovery from long runs:
- Is the athlete eating enough during the long run?
- Is the athlete eating enough in general or are they struggling with low energy availability (intentional or unintentional)?
- What was the intensity of the run?
- Did the athlete run on a hard or soft surface?
- Was footwear cushioned enough?
- How well are they sleeping?
- How often are they doing these long runs? Are they exceeding three hours weekly or was it an uncommon occasion in peak training?
- How were these long runs periodized?
In coaching practice, I have had to weigh injury risk and training readiness often. I have some athletes for whom, based on injury risk, I limit long runs to 16-18 miles (3-3.5 hours). For others, I’ve had them do 20-mile-long runs that take approximately 3.5 hours – and in two of those scenarios, resulted in a sub-4 marathon without bonking for their first marathon.
Weighing Risk vs Reward
Every training decision is a balance of risk vs reward. Injury risk should be minimized, but not at the expense of compromising race readiness. In a scenario where a runner will exceed 3 hours to cover 25km/15 miles, I’d much rather they run for longer than 3 hours and get at least one 16-mile run before race day.
This risk vs reward scenario applies to every single athlete. If I observe an athlete who recovers slowly from long runs, I adjust their long run duration accordingly. If an athlete is remarkably fatigue resistant, has adequate energy availability, and fuels well on their long runs, then they may go past 3 hours in training if it is appropriate for their training.
Coach Humphry gives a prime example of weighing the risk vs reward of long runs. In a Youtube video, he rationally explains how to approach to training for slower marathoners focused on finishing. “Since you are overall just trying to build your general endurance, I definitely have more leeway on that three hour rule. If you have to bump that up to 3:15 or three and half hours…that’s probably okay…because the amount of work you are trying to recover from is less from an intensity standpoint [during the week].”
Bioenergetics Are Not the Only Factor You Train
If only exercise science was as easy as straight bioenergetics. Bioenergetics is the easy part! However, endurance training and performance are much more complex: biomechanics, the endocrine system, the nervous system, the digestive system, and more all come into play.
When I train ultra runners, I do not cap long runs at 3 hours. These are runners who will be on their feet for 6-8 hours (sometimes much, much more). They need to know what their gear feels like, how to fuel, what mental tactics to use, etc.
As the Science of Ultra points out, the diminishing returns of bioenergetics adaptations do not negate the occasional use of a long run past three hours in duration. When used sparingly, a long run beyond three hours can improve fatigue resistance and allow precise calibrations in fluid and fuel intake. For many runners also, it provides a significant psychological boost. You aren’t just a pair of legs and lungs running a marathon; your brain plays a role in fatigue also.
Base vs. Specific Training Affects Long Run Duration
There is a significant difference between the long runs you do when training for a race versus when in a base phase. I generally do not prescribe long runs over two hours when in non-specific training. None of my athletes are doing 20 mile runs more than 12 weeks out from a marathon or ultra.
Ultra running expert Jason Koop stated that long runs should not be under three hours outside of specific race training. However, when training for a race, he extends long runs to four hours or beyond. Top trail running coach David Roche takes a similar approach, limiting long runs to 2.5 hours except in the 3 months prior to a goal race. Then, long runs exceed three hours because their purpose shifts to ensuring athlete readiness for the distance.
Yes, those are expert opinions – just like how the three hour rule is an expert opinion. One expert opinion does not automatically outweigh others. These approaches warrant equal consideration to the three-hour rule.
Race preparation training is different from base training, due to the principle of specificity. Think of your quality sessions: you do not do the same 10-mile marathon pace run you do in peak marathon training during earlier phases of training. If periodization applies to training intensity and training volume, logically then it also applies to long run duration.
Arguably, the 3-hour long run rule can take a periodized approach. Keep your long runs to 25-35% of your weekly distance or less than 2.5 hours when not in specific training. Once you are closer to your race (8-12 weeks prior), periodize your training to prioritize race readiness, even if you have a 3.5-hour long run or a long run that is 40% of your weekly mileage.
The Long Run Matters, but Not at the Expense of Weekly Volume
That all said, your long run should not supersede the overall weekly training volume. As noted above, weekly volume is also predictive of marathon performance.
If you have dead legs on most of your long runs, you may be running too much or too hard during the rest of the week. It’s a fine balance. If you find yourself struggling to complete the follow week’s training (and your sports nutrition is adequate), it may be worth taking limiting the duration of your long run. Alternatively, you may perform a limited number of runs over three hours, and follow those with cutback weeks to ensure recovery.
Conclusion: Evidence Does Not Support Always Limiting Long Runs at Three Hours
The three-hour rule is not definitive. It may be appropriate in some scenarios and inappropriate in other scenarios. The three-hour cap on long runs has merit for injury-prone marathoners. However, it’s not an absolute rule nor is it applicable in every scenario. Evidence-based, individualized coaching means that some runners will exceed three hours in a long run in specific training – and others will barely approach it. If it takes you 3.5 hours to complete 20 miles in training, do not let this cause you significant stress.
And this is not to say that coaches abiding by the three hour long run rule are wrong. This is simply to do what we do in exercise science: examine the evidence, assess practice, and debate. You see similar things with training intensity distribution: coaching methodologies establish it as primary practice, while the scientists themselves are still in heated debates years later.
If I were to design a guideline around long run duration it would be: “The exact distance should based on the goal race, the athlete’s durability and injury risk, and recovery rate. Intra-run carbohydrate intake should be high, walk intervals and softer surfaces may be encouraged for injury-prone athletes, and the following days of programming should ensure adequate recovery. Periodize long runs on a macro and micro scale.”
Your long run is not the biggest effort of the training cycle: the race is. However, the long run should ensure readiness for that race, without leaving your race in training.
And when you are looking at training guidelines, if someone says “studies say,” politely ask them to cite which studies.
References Not Linked
Daniels, J. (2022). Daniels’ Running Formula. (4th ed.) Human Kinetics.
Humphrey, L., Hanson, K., & Hanson, K. (2012). Hansons marathon method. Velo Press
Jeukendrup, A. & Gleeson, M. (2016). Sport nutrition. (3rd ed.) Human Kinetics.