Best Marathon Shoes for Women: Training Pairs and Race-Day Picks

Our pick is the New Balance 1080 v15: the deepest, most even cushioning of the neutral trainers here, in a women's size run that goes both narrower and wider than most brands offer. For the longest runs, the Saucony Triumph 24. If you want support, the Brooks Ariel GTS 26, which is a women's-specific model rather than a men's shoe in different colors. On race day, the Saucony Endorphin Pro 5.

The short answer

What the research actually says about women's running shoes

"Women's" on a shoe box can mean two very different things: a genuinely different last, or the men's shoe graded down and recolored. Two published studies are worth knowing about.

Wunderlich and Cavanagh (2001), in Medicine & Science in Sports & Exercise, measured foot and leg dimensions in a large young sample. Men had longer and broader feet for a given stature, and after normalizing to foot length the sexes still differed significantly in two calf, five ankle and four foot-shape variables. Their conclusion, in their words: female feet and legs are not simply scaled-down versions of male feet but differ in shape, particularly at the arch, the lateral side of the foot, the first toe and the ball of the foot — and those differences should be taken into account in the design and manufacture of women's sport shoes.

Krauss and colleagues (2010), in Research in Sports Medicine, scanned 424 feet and four men's running shoe lasts in US women's sizes 6.0-9.5 — lasts that are also used to manufacture women's shoes. Differences in width measures between lasts and foot types varied substantially, by 0 to 9 mm. Length grading matched (under 1 mm), but width grading was larger in the lasts than in real feet, by 3.5 to 5.9 mm. Their conclusion: the use of down-graded men's lasts for women's shoes has to be questioned.

What that means when you are shopping, and what it does not mean:

• It does mean fit is the variable to optimize. If a shoe is graded from a men's last, the width may be off by several millimeters in your size, and width error is felt as heel slip, forefoot squeeze or a hot spot at mile 16.
• It does mean women's-specific models are worth a look. The Brooks Ariel GTS 26 exists as a women's counterpart to the men's Beast, not as the same shoe in pink.
• It does not mean women need different technology. Nothing in either study says women need more cushioning, less drop or a softer midsole. Those are individual preferences.

Women's width letters are not men's width letters

This trips up more marathon shoppers than any midsole. The same letter means a different measurement depending on the size run:

So a women's D is a wide, and a men's D is a standard. If you have ever bought a men's shoe in your converted size because the color was better, that is why it felt roomy. Brooks labels the Ariel GTS 26 as Medium (1B), Wide (1D) and Extra Wide (2E) for exactly this reason. Our wide feet page goes through the letters in full.

The training picks, side by side

One honest caveat about every weight figure you will read online, here included: labs weigh a men's sample size, usually a US 9. Your women's shoe will weigh less. Use the numbers to compare shoes with each other, not to predict what is on your foot.

A two-pair marathon rotation

Buy the second pair a few weeks after the first so they do not both wear out together. Our mileage and rotation guide has the arithmetic.

Race day: plated or not

Carbon racers are worth a clear-eyed look rather than a leap. World Athletics caps road stack height at 40 mm and both racers here sit near it, which is part of why they feel unstable at first. Two rules we would hold to:

• Run long in it before you race in it. A 20-minute test does not tell you how a plated shoe treats your calves at mile 22.
• Pick the steadier one for a first marathon. That is the Endorphin Pro 5, with its 8 mm drop and forked plate, rather than the lightest, springiest option.

If plated shoes do not suit you, that is a perfectly good answer — see best marathon shoes without a carbon plate. And if you want support on race day, there is no supportive super shoe on the market; race in the Ariel GTS 26 and accept the weight.

How to choose

1. Get the width right first. Remember that a women's B is standard and a women's D is wide. Krauss 2010 found width errors of up to 9 mm between men's-derived lasts and real women's feet; that is the error you are trying to avoid.
2. Buy for the long run, then for the midweek runs. The long-run shoe is the one that decides whether the block goes well.
3. Leave toe room. The AAOS asks for at least half an inch beyond your longest toe, and Cleveland Clinic's fitting guide for a thumb's width, with a snug heel. Feet swell over 26.2 miles.
4. Only buy support if you want support. Ryan and colleagues (2011) assigned women runners shoes by arch type during half-marathon training and the prescription did not reduce pain. Comfort beats a category label.
5. Decide about race shoes by week 10, not race week, so you have time to run long in whatever you choose.

Common mistakes

• Buying the men's version in a converted size. Different last, different width standard, usually a roomier heel than you want.
• Assuming a women's shoe means a softer shoe. Nothing in the research supports that. Choose your ride, not your marketing segment.
• Racing 26.2 in shoes you have run 6 miles in. The most common and most painful mistake on this page.
• Running the whole 18 weeks in one pair. At about 500 miles the cushioning is done, and a marathon build gets there.
• Ignoring heel slip. A narrower heel is one of the shape differences Wunderlich and Cavanagh identified. Try a runner's loop before you change model.