How to choose running shoes (without the gait-analysis dogma)

Here is the whole guide in one sentence: buy the pair that feels most comfortable on your foot, in your normal running, and ignore almost everything else. The "stand on this treadmill, we'll analyse your gait and prescribe a pronation-control shoe" routine sounds scientific, but the evidence under it is thin. What does have support is much less impressive to watch: your own comfort.

Start with comfort, because that's the part with evidence

The standout finding in running-shoe research isn't about arches or plates. It's that runners who pick shoes they find comfortable tend to get hurt less. In one analysis of 527 leisure-time runners, those who perceived their shoes as more cushioned and rated them more highly overall had substantially lower injury risk. The idea isn't new: earlier work on self-selected, comfort-based footwear pointed the same way, with comfort-chosen inserts linked to fewer lower-limb injuries than assigned ones.

Researchers call this the "comfort filter": you intuitively choose footwear that keeps you in your natural movement pattern, and that seems to protect you. It's not glamorous. It's also the most defensible rule going.

The pronation sales pitch, and why it doesn't hold up

Walk into many running stores and you'll be filmed jogging, told you "overpronate", and steered toward a stability or motion-control shoe. The problem is that prescribing shoes by foot type or pronation has never been shown to prevent injury.

A 2009 systematic review put it bluntly: searching for controlled trials of cushioned, pronation-control shoes individualised to foot type, the authors found no original research supporting the practice. It is not evidence-based. Later work agreed. A prospective study of nearly 1,000 novice runners found that foot pronation was not associated with increased injury risk when people ran in a plain neutral shoe.

So what is the gait analysis actually doing? Mostly changing how you feel. In a 2025 single-blind trial, 21 women ran in identical shoes labelled either "basic" or "gait-matched". They rated the supposedly matched pair about 17mm higher on comfort and around 20mm higher on perceived injury reduction, despite no meaningful biomechanical difference between the shoes. That's a marketing placebo. It can make you happier with a purchase. It does not lower your injury risk, and the study never claimed it did.

The prescription of distance-running shoes by foot type is not evidence-based.British Journal of Sports Medicine, 2009

If you overpronate, you do not automatically need a "control" shoe. You need a shoe that feels good when you run.

How a running shoe should actually fit

Fit is where comfort gets concrete. Aim for roughly a thumb's width, about half an inch or 1cm, between your longest toe and the front of the shoe. The heel and midfoot should feel snug and locked, while your toes can wiggle freely. Snug back, room up front.

Try shoes on at the end of the day or right after a run. Feet swell as the day goes on and as you run, so a pair fitted first thing in the morning can feel a size too small by your evening session.

And forget the idea of a break-in period. A running shoe that pinches, rubs or aches in the shop is the wrong shoe. The whole point of the comfort filter is that the right pair feels right immediately. If it hurts on day one, put it back.

Cushioning, ground feel and heel drop: what they change

These are preferences, not prescriptions. More cushioning gives a softer, more isolated ride; more ground feel gives a firmer, more connected one. What cushioning does not reliably do is protect your joints by lowering impact forces. The evidence that softer midsoles reduce ground reaction forces is weak, partly because runners unconsciously adjust their stride to whatever's underfoot. Max-cushion shoes can be wonderful to run in. Just buy them because you like the feel, not because you've been promised impact protection.

Heel-to-toe drop, the height difference between heel and forefoot, works the same way. A randomised controlled trial found that drop didn't change overall injury risk. It shifted where load went: lower drop nudged stress toward the ankle, foot and Achilles, higher drop toward the knee. There is no single correct number. If your knees are cranky, a slightly higher drop may suit you; if your Achilles complains, the opposite. Choose for your body, and change drop gradually.

Daily trainer, race shoe, trail shoe: do you need all three?

Most runners need one. A daily trainer is the cushioned, durable, do-everything shoe you put easy kilometres on. If you only buy one pair, buy this.

A race-day shoe is lighter, firmer and often carbon-plated, built for going fast on race day. A trail shoe adds grippy lugs and protection for soft, uneven or rocky ground, and it earns its place only if you actually run off-road. If all your running is on pavement and you're not racing, the daily trainer covers you completely.

Carbon plates, honestly

Super shoes are real, but oversold. Pooling 13 studies, carbon-plated shoes improve running economy by roughly 2-3% on average, worth on the order of one to two minutes over a marathon. Useful if you're chasing a time. Largely irrelevant if you're not.

Two honest caveats. First, the benefit is speed-dependent and individual. It shrinks at easy paces, on the order of 1.4% at around 3:30-marathon pace, and the stiff plate tends to help least on slow, uphill or technical running. A 2023 meta-analysis found responses ranging from about 9.7% faster to 1.1% slower depending on the runner and the model. Some people simply don't click with them.

Second, they carry risk. Carbon-plate footwear has been linked to bone stress injuries, including high-risk navicular stress injuries, because the stiff plate creates loading patterns your foot isn't used to. Treat super shoes as an occasional fast-day tool, not your daily trainer.

Rotation, mileage and when to replace

Owning one perfect pair feels frugal, but rotating shoes is the smarter move. Among 264 recreational runners over 22 weeks, those who rotated multiple pairs had a 39% lower injury risk than single-pair runners, probably because different shoes vary the load on your tissues.

Replace within the familiar 480-800 km window. That's not arbitrary: midsole foam loses around 20% of its shock absorption by about 480 km and up to 30% by 800 km, even when the shoe still looks fine. The foam dies before the outsole wears through, and heavier runners burn through it faster. Don't run a shoe until the tread is gone.

A no-upsell checklist for your first pair

Decline the gait analysis as a prescription; treat it, at most, as free entertainment. Try several pairs and pick the most comfortable. Check for a thumb of toe room, a snug heel and no break-in pain. Shop late in the day. Don't pay extra for pronation control or a carbon plate you don't need. Buy a forgiving daily trainer, add a second pair to rotate when you're running regularly, and replace around 480-800km. That's it.

FAQ

What is the most evidence-backed way to pick running shoes?

Choose the pair that feels most comfortable during your normal running. Comfort-chosen footwear is linked to lower injury risk in leisure runners.

Do I need gait analysis and a stability shoe?

Prescribing shoes by pronation has not been shown to prevent injury. Overpronating does not automatically mean you need a control shoe.

How should running shoes fit?

Leave toe room, lock the heel and reject any rubbing. Fit turns comfort into something you can check before you buy.

Do I need carbon-plated shoes for daily training?

No. Save super shoes for races and key speed sessions if you respond well to them. Most runners need one reliable daily trainer first.

Does heel drop prevent injury?

Drop and cushioning shift load rather than magically remove it. Change drop gradually if you switch models.

Sources

• Is your prescription of distance running shoes evidence-based? (BJSM, 2009)
• Foot pronation is not associated with increased injury risk in novice runners wearing a neutral shoe (Nielsen et al., BJSM 2014)
• Running injury paradigms and footwear design (Frontiers focused review, 2022)
• Mythbusting and two new paradigms: preferred movement path and comfort filter (Nigg et al., 2015)
• Shoe cushioning perception, comfort and injury risk (secondary analysis, 2025)
• Gait-matched shoe recommendations and perception: single-blind RCT (2025)
• How running shoes should fit (REI Expert Advice)
• Heel-to-toe drop and injury risk RCT (Malisoux et al., 2016)
• Metabolic effects of carbon-plated running shoes: systematic review and meta-analysis (2025)
• Carbon-plate running economy at different speeds and on trails (RunnersConnect)
• Individual variation in super-shoe response: 2023 meta-analysis (Marathon Handbook)
• Bone stress injuries in runners using carbon fiber plate footwear (2023)
• Can parallel use of different running shoes decrease injury risk? (2015)
• How long do running shoes last (Marathon Handbook)