Running shoe advice is full of rules about arch types, overpronation and drop. We read the trials and the sports-medicine guidance to separate what holds up from what is habit.
The Review NYU is not affiliated with New York University.
We do not currently earn a commission from any link on this page. If links ever become affiliate links, we will label them clearly.
Short answer. Start with comfort and fit, not foot-type labels. Large military trials found that matching shoes to arch shape did little to reduce injuries, and a 2015 review in the British Journal of Sports Medicine argues runners should use their own "comfort filter." Get both feet measured, leave about a thumb's width of toe room, make sure the heel does not slip, and pick the most comfortable shoe from a short jog. Some features may matter for some runners: in trials, softer shoes helped lighter runners and motion-control shoes helped runners with pronated feet. If you run regularly, avoid big jumps in heel-to-toe drop.
We did not test any shoes hands-on, run a gait lab or fit anyone in a store. This is a research guide. We read the guidance published by the American Academy of Podiatric Sports Medicine (AAPSM) and the American Podiatric Medical Association (APMA), shoe makers' own definitions (Brooks), a retailer's explainer (Fleet Feet), RunRepeat's lab-measured drop data, and six peer-reviewed studies on shoes and running injuries, all checked 12 November 2026.
We gave the most weight to randomized trials and large prospective studies, less to expert opinion, and the least to marketing language. Where sources disagree, we say so rather than pick a winner. Nothing here is medical advice: if you have pain, a past injury that keeps coming back, or a foot condition, a sports-medicine clinician who can examine you will know more than any article.
Much of the shoe-buying advice you hear in stores rests on two older ideas: that softer shoes prevent injury by reducing impact, and that "overpronators" need stability shoes. The research is messier than that.
A widely cited 2015 review in the British Journal of Sports Medicine by Benno Nigg and colleagues at the University of Calgary concluded there was "no conclusive evidence" that vertical impact forces are associated with running injury, and "no evidence that foot pronation (eversion) is a variable responsible for running injuries." The authors also noted that reported injury rates in studies ranged from about 15% to 85% of runners, with no obvious change over 40 years of shoe development. Their proposal, the "comfort filter," is that runners naturally pick comfortable products and that doing so may itself lower injury risk. As supporting evidence they cite an earlier military study in which soldiers who chose their most comfortable of six insoles had 53% fewer lower-extremity injuries than a control group. That is one study of insoles, not shoes, so treat it as suggestive.
Later randomized trials from the Luxembourg Institute of Health found that some features do seem to matter, for some runners. In a 2020 trial of 848 recreational runners, those given a harder version of the same shoe had a higher injury risk than those given a softer version (subhazard ratio 1.52). In a 2016 trial of 372 runners, motion-control shoes were linked to lower injury risk overall, an effect seen only in runners with pronated feet. Both trials were partly run with Decathlon, a shoe maker, which is worth knowing.
The honest summary: comfort is a reasonable first filter, specific features may help specific groups, and no single shoe category has been shown to prevent injuries for everyone. Training load, meaning how much and how fast you increase mileage, is something AAPSM also flags as a major factor.
The AAPSM calls the midsole "the most important part of the shoe in terms of cushioning and stability." It is usually EVA or polyurethane foam, sometimes with brand-specific additions. Two terms get confused: stack height is how much material sits between your foot and the ground, while drop (next section) is the difference between heel and forefoot height. A tall shoe can still have a low drop.
More cushioning is not automatically better. In the 848-runner trial, the protective effect of the softer shoe showed up only in lighter runners (subhazard ratio 1.80 for hard shoes in the lighter half); heavier runners showed no significant difference. Body mass was split at the median, which was 78.2 kg (about 172 lb) for men and 62.8 kg (about 138 lb) for women in that study. That cuts against the common assumption that heavier runners need the most cushioning.
| Category | Stack / cushioning | Drop | Notes from AAPSM |
|---|---|---|---|
| Barefoot | Very little to no midsole | Zero | Very flexible and light; marketed to a subset of forefoot strikers |
| Minimalist | Low | Low | Light, more ground feel; similar to a racing flat |
| Traditional (neutral, stability, motion control) | Moderate | High | Split by how much added support they have |
| Maximalist | High | Low | Often a forefoot rocker; popular with runners who want more shock absorption |
| Carbon-plated racer | Usually high | Varies | Very stiff; AAPSM says these are designed for racing, not as everyone's daily trainer |
Brooks, a shoe maker, frames the max-cushion versus low-profile choice as mostly personal: "ultimately it's what you, the runner, prefer." Given the evidence above, that is not a bad stance.
Drop (also called offset) is the height difference between the heel and the forefoot, in millimeters. There is no single standard for what counts as "low" or "high," as the table below shows.
| Label | AAPSM | RunRepeat (lab categories) |
|---|---|---|
| Zero | Included in low (0-6 mm) | 0 mm (up to 0.9 mm measured) |
| Low | 0-6 mm | 1-4 mm |
| Mid | Not a separate category | 5-8 mm |
| High | Typically 7-12 mm | 8+ mm |
Fleet Feet, a running retailer, says HOKA shoes usually sit around 4-5 mm, more traditional Brooks and ASICS models around 8-10 mm, and Altra is known for zero drop. Its physical therapist source says the research "isn't conclusive enough" to say whether drop affects injury rates.
The best evidence we found is a 2016 randomized trial of 553 leisure-time runners given otherwise similar shoes with 10, 6 or 0 mm drop and followed for six months. Overall, drop did not change injury risk. In a secondary analysis, lower-drop shoes were linked to lower risk in occasional runners (hazard ratio 0.48) but higher risk in regular runners (hazard ratio 1.67). Secondary analyses are less reliable than the main result, but the practical reading is sensible: if you have run regularly in a high-drop shoe, do not jump straight to a much lower one.
Check the drop number with some skepticism. RunRepeat, which cuts shoes in half and measures them following World Athletics rules, reports that brand-stated drops are often off, in some cases by several millimeters. For example, it lists one New Balance model at 8 mm stated versus 3.3 mm measured.
Fit is where expert sources agree most. Their guidance, summarized:
AAPSM also suggests wearing a shoe for at least 10 minutes in the store and, if allowed, jogging briefly outside. Return policies for worn shoes vary by store and brand, so read the terms before you run outside in a new pair.
Pronation is the normal inward roll of the foot after landing. Stores have long sorted runners into neutral feet, overpronators and supinators, then matched them to neutral, stability or motion-control shoes. Brooks still describes its support shoes as "ideal" for runners with "mild to severe overpronation with low or flat arches," using medial posts or its GuideRails system. The APMA says shoes should match your arch type. Those are the maker's and the association's positions; the trial evidence is more mixed.
Three US military trials (Army, Air Force and Marine Corps, about 7,200 recruits in total) assigned shoes based on foot shape: motion control for low arches, stability for medium, cushioned for high. Compared with recruits who all got a stability shoe, there was little difference in injury rates (pooled rate ratio 0.97 for both men and women). In a one-year Danish study of 927 novice runners all wearing neutral shoes, pronated feet were not at higher risk than neutral feet; pronators actually had slightly fewer injuries per 1,000 km. The authors said more work is needed on highly pronated feet, which were rare in the sample (18 feet).
The 2016 Luxembourg trial found the opposite pattern among regular recreational runners: pronated-foot runners in standard shoes had higher injury risk than neutral-foot runners, and motion-control shoes reduced risk in the pronated group (hazard ratio 0.34, from a subgroup of 94 runners). Small subgroups produce wide uncertainty.
Our reading: a gait analysis or foot-type label is not a reliable prescription by itself. If you have a clearly pronated foot and a history of problems, a stability shoe is a reasonable thing to try, ideally with input from a clinician. If you feel fine in a neutral shoe, there is little evidence you need to change. AAPSM's own list of factors for choosing goes well beyond foot type: running surface, injury history, body mass, experience, training volume and race goals.
Expert guidance on lifespan conflicts. AAPSM says most running shoes last 350-500 miles; APMA says to replace them after 600-800 miles or every 6-8 months. Neither cites a study for its figure. Logging mileage and checking for a worn-through outsole, a tilted heel counter or a midsole that feels flat is more useful than any single number.
Specs on shoe boxes and websites can be marketing, not measurement; our guide on how to read a product spec sheet covers what to trust. And before you put miles on a pair, see how to test a product during the return window.
Not necessarily. Military trials found that assigning shoes by arch shape did little to reduce injuries, and a Danish study found pronated feet were not at higher risk in neutral shoes, though one Luxembourg trial found motion-control shoes helped runners with pronated feet. If a neutral shoe feels fine, there is little evidence you need to switch.
About a half to a full thumb's width, or roughly an index finger's width, between your longest toe and the end of the shoe, according to the American Academy of Podiatric Sports Medicine and Brooks. Measure on your longer foot while standing.
A large trial found drop did not change overall injury risk, but lower-drop shoes were linked to more injuries in regular runners and fewer in occasional runners. If you already run regularly, stay close to the drop you are used to and change gradually.
The evidence does not clearly support that. In a trial of 848 runners, softer shoes lowered injury risk only among lighter runners, with no significant difference for heavier runners.
Guidance varies: AAPSM says most shoes last 350 to 500 miles, while APMA says 600 to 800 miles or every 6 to 8 months. Track your mileage and look for a worn outsole, tilted heel counter or flattened midsole.