Tart Cherry Juice for Runners: Recovery and Race Day

Tart cherry juice is sold on a promise it does not keep. The claim is that it makes you less sore. Pooled across 19 trials in trained athletes, it does not: "no significant pooled effects were found for muscle soreness, CK, IL-6, TNF-alpha, ROM, or most subgroup time points."

What the same analysis did find is arguably more useful. Your ability to produce force comes back faster, at every time point measured. That is a different product than the one on the label, and it changes when you should bother drinking it.

What it actually does

A 2026 meta-analysis (Daab et al., Sports Medicine Open) pooled 19 trials in trained athletes under PRISMA 2020. Maximal voluntary contraction recovery improved across all time points: effect sizes of 0.63 immediately post-exercise, 1.12 at 24 hours, 1.29 at 48 hours, 2.14 at 72 hours and 4.82 at 96 hours. Countermovement jump improved at one point only, 48 hours (ES 1.41). C-reactive protein was reduced post-exercise and up to 48 hours (ES -0.46, -0.73 and -0.68).

Now the caveats, which the authors state themselves. Heterogeneity was substantial (I-squared 69-93%). "Certainty of evidence ranged from very low to moderate." The conclusion sentence reads: "findings are heterogeneous and supported by low-to-moderate certainty of evidence, warranting cautious interpretation." A sensitivity analysis adds that "statistical significance at selected time points (MVC post and 72 h; CMJ 48 h; CRP 24-48 h) was influenced by individual studies" — remove one trial and some of those results move.

That is a real finding with a wide error bar around it, not a settled fact. Anyone selling you certainty here is selling something.

The marathon study, and the protocol that came out of it

The anchor trial for runners recruited 20 recreational marathon runners (Howatson et al., 2010). They took tart cherry juice or placebo "for 5 days before, the day of and for 48 h following a Marathon run" — eight days in total.

Isometric strength recovered significantly faster in the cherry group (P=0.024). The next sentence of the abstract is the one nobody quotes: "No other damage indices were significantly different." The paper counts muscle soreness among the damage markers it measured, so that sentence covers soreness and creatine kinase — and the 19-trial pooled analysis later reached the same null independently.

Inflammation markers moved: C-reactive protein (P<0.01) and uric acid (P<0.05). Interleukin-6 was also significant in this single trial (P<0.001), but the later pooled analysis found no significant effect for IL-6, so treat that one as a single-study result that did not replicate. Total antioxidant status ran about 10% higher at every post-supplement measurement (P<0.05), and lipid peroxidation was lower at 48 hours.

On dose, be careful with what is actually documented. The eight-day timing is verbatim from the abstract. The per-day amount is not: the full text is not publicly retrievable, and the reported protocol comes from secondary sources — approximately 8 fl oz (about 237 mL) of bottled ready-to-drink juice, twice a day, roughly 120 cherries' worth per day. It was not a concentrate to be diluted, which matters, because buying a syrup-style concentrate and drinking 237 mL of it is a completely different dose.

The reversal: before a race, and only if you are well trained

Standard advice splits these two supplements neatly: beetroot before, cherry after. Two independent sources say that split is wrong, and the correction runs in the counter-intuitive direction.

A crossover trial gave Montmorency cherry concentrate 30, 90 or 150 minutes before a 15 km time trial, with 20 participants split into 10 recreationally active and 10 trained. Result, verbatim: "MCC significantly improved TT performance, but not exercise economy. The greatest improvement in performance occurred following MCC90min compared to US (US 1603.1 ± 248 s vs. MCC90min 1554.8 ± 226.7 s, 2.83% performance improvement). Performance was significantly enhanced for trained (US 1496.6 ± 173.1 s vs. MCC90min 1466.8 ± 157.6 s) but not recreationally active participants." The authors add: "The ergogenic effect was greater for trained participants."

Two limits before you act on it. That was a cycling time trial, not a run, so it counts as evidence from a comparable endurance sport rather than direct evidence for runners. And it is one trial with ten people per group.

The mechanism is worth knowing because it explains the split. Plasma phenolic metabolites (ferulic and vanillic acid) peaked around 90 minutes after intake, while plasma nitrate and nitrite did not change. This is the polyphenol pathway, not the nitrate pathway — a different lever from beetroot entirely.

A 118-study meta-analysis then points the same way. "Trivial but significant benefits were demonstrated for consumption of nitrate and polyphenol-rich foods (SMD=0.15 and 0.17, respectively, p<0.001)", with Montmorency cherry named among the polyphenol foods showing benefit. The key line: "Well-trained males (VO2max 65 ml.kg.min-1 or above) exhibited small, significant benefits following polyphenol, but not nitrate consumption." Its conclusion: "Highly trained endurance athletes do not appear to benefit from consuming nitrate-rich foods but may benefit from polyphenol consumption."

Read the row labels carefully, because this is the opposite of how the two products are usually sold: the better trained you are, the less beetroot does and the more likely tart cherry is to do something. Both effects are small — the authors of these analyses use the words "trivial" and "objectively small" themselves. The beetroot half of the picture, including the dose that actually reaches the studied threshold, is in the beetroot juice guide.

One scope note on the female row: it comes from the performance analyses. The recovery meta-analysis reported on trained athletes without a sex breakdown, so it says nothing either way about women and post-race strength recovery.

How to read a tart cherry label

Try to compare two tart cherry products on dose and you will discover the real problem: this category currently has no agreed unit of measure that lets a consumer compare products side by side. That is a statement about the category, not an accusation against any company.

The two halves of the supplement shelf arrive there for different reasons, and merging them weakens the argument. On the beetroot side the cause is raw material variation — nitrate content tracks soil and storage. On the tart cherry side there are two causes stacked on top of each other.

• The raw material varies enormously. In one study, the anthocyanin concentration of different batches of fresh sweet cherries "varied almost 20-fold". That is sweet-cherry data, cited here as evidence that cherry raw material swings wildly from batch to batch rather than as a measurement of tart cherry products. Across cultivars, total phenolics span 74-754 mg per 100 g fresh weight and anthocyanins span 21-285 mg.
• The measurement method is not standardised. The number a laboratory reports depends in part on which assay it used, so even two honestly measured products may not be on the same scale.

This is not an outsider's complaint. The authors of a 29-study review of cherry research wrote it into their own conclusion: "Developments of stable and standardized cherry products that retain nutrient composition of fresh cherries and of placebos devoid of polyphenols are desperately needed to precisely assess the health promoting effects of cherry consumption. It is important that all intervention studies report at least the daily total amounts of phenolics and anthocyanins served to study participants." Researchers are asking for the same numbers you are.

What brands print instead

Across the brands checked in August 2026, none stated anthocyanin milligrams on their product pages, ingredient pages or official dosage articles. What sits in that space is a substitute unit.

• Cherry counts: "60 TART CHERRIES / 8 OZ" and "50 TART CHERRIES / 8 OZ" on two products from one brand; "approximately 3,000 Montmorency cherries" per 946 mL from another; "Approximately 1,600 cherries per bottle" from a third.
• Weight of fruit: "made from 3 lbs. of organic cherries."
• Equivalences rather than contents: one bottle "delivers the antioxidant power of 1,000 tart cherries" and has "five times the antioxidants of sweet cherries"; on the beetroot side of the same shelf, "1 scoop = 40X the functional benefits of fresh beets."

Every one of these shares a defect: the denominator cannot be falsified. There is no legal or standardised anthocyanin content for "one cherry", so no count can be converted into a dose, and no two counts can be compared.

Two pieces of evidence show that cleanly. Within one brand, a 59 mL recovery shot and a 74 mL concentrate pouch are identical on every nutrient both labels declare — 100 kcal, 23 g carbohydrate, 20 g sugar, 1 g protein, 45 mg sodium, 320 mg potassium — while claiming "60+" and "40" cherries respectively. Same measurable composition, a 50% gap in the claim. Both cannot be right.

Across brands it is worse, because the numbers are not even the same kind of claim. Two of them are ingredient statements: "approximately 3,000 Montmorency cherries" in one bottle, "Approximately 1,600 cherries per bottle" in another. The third is an equivalence — a bottle that "delivers the antioxidant power of 1,000 tart cherries" is not telling you that 1,000 cherries went into it. Set 3,000, 1,600 and 1,000 side by side and you are holding two quantities up against an analogy. These figures do not qualify to sit in the same column.

The single most telling detail

On one brand's official article about how much tart cherry juice to drink per day, the only milligram figure on the entire page is "3mg of added melatonin". Anthocyanins get a qualitative sentence instead. The same company lists "320 mg of magnesium glycinate" on another product.

So precise milligram labelling is clearly within reach when the ingredient is added rather than inherent. The molecule the recovery claim rests on is the one without a number. That is a verifiable contrast, stated as such — draw your own conclusion from it.

The category has a working counter-example

This is not a story about brands being untrustworthy; it is a story about knowing what to look for. On the nitrate side there is a product where the number checks out: a 70 mL sports shot claiming 400 mg of nitrate was independently measured at 6.41 ± 0.60 mmol, which is about 397 mg. Claim and laboratory assay agree.

That is what a checkable product looks like, and it exists. On the tart cherry side there is currently no equivalent, because there is no published number to check against.

Can you check a serving against the studied amount?

The exercise studies in that 29-study review served the equivalent of 50 to 270 cherries per day, which the review's own authors describe as a high dose. The obvious next move is to hold a product's serving up against that band. On the cherry side you cannot, for the reason just given: some of the counts printed on labels are equivalence claims rather than statements of fruit content, and setting one of those against an amount of fruit people actually ate would repeat the exact error this section is about.

The beetroot side shows what the check looks like when the arithmetic is available. Of the six beetroot powders in a 2019 independent testing round, not one stated a nitrate figure on its label and not one delivered 5 mmol in a single serving; measured contents ran from 0.43 to 2.56 mmol, so the weakest would take fourteen scoops or more to reach the research dose. Neither category is misstating a figure. Neither publishes one.

Two cautions on the 50-270 band before you use it as a target. Of the 29 human studies in that review, 20 used tart cherries, 7 used sweet cherries and 2 were unspecified, so it is a mixed pool. And the counts still cannot be converted to milligrams for the reasons above.

"Concentrate" means three different things

• True syrup-style concentrate that must be diluted, typically served as a 30 mL or 40 mL dose. One brand instructs a 1-to-5 dilution with water, seltzer or juice.
• Products named "concentrate" that are drunk neat — one 2.5 fl oz (74 mL) pouch is simply thicker than the same brand's ready-to-drink bottle, with no dilution step at all.
• Ready-to-drink juice, which is what the marathon trial used.

Confusing the first two in either direction ruins the dose. Dilute the pouch and you drink a fraction of what you intended; drink 237 mL of the syrup and you are far past any studied protocol.

The sugar you are also buying

Nutrition panels are far more stable over time than prices, so this is the part worth planning around. Running the eight-day marathon protocol on one common ready-to-drink product means about 200 kcal and 50 g of sugar per day for eight days; on the same brand's original formula, 240 kcal and 58 g. Pushing to the 270-cherry top of the studied band would mean roughly 4.5 bottles a day: about 450 kcal and 112.5 g of sugar.

Whether that is a cost or a bonus depends on where it lands. During a carb-loading week it may be doing useful work; in a base block it is a large daily sugar addition with no training purpose. Set it against your actual carbohydrate targets in the daily nutrition guide for runners.

One thing that number does not automatically imply. A dose-response meta-analysis of 12 randomised trials, graded with GRADE, concluded that "tart cherry supplementation did not have significant effects on anthropometric and glycemic indices", while finding improvements in lipid profile. So the sugar is real arithmetic, but the knock-on to glucose measures you might expect has not appeared in the pooled trials. No trial has been run in people with diabetes at all, so if you manage blood sugar, the 50 g figure is information to take to your doctor rather than a decision this page can make.

On price, only structural conclusions age well. Within a single brand, covering one race week can differ by roughly three-fold depending only on which pack size you buy. And bulk is not automatically cheaper: the multi-serve bottle came out about 18% more expensive per serving than the single-serve pouches.

When it does not work

A trial with 36 participants took 2 x 250 mL daily for 9 days and performed heavy eccentric elbow flexor exercise on day 5. The damage was real — maximal voluntary contraction fell by up to 26.8% and creatine kinase rose as high as 1385 U/L — but there were no differences between groups. The conclusion: "neither TC nor POM enhance recovery from high-force eccentric exercise of the elbow flexors."

So the honest boundary is narrower than the marketing. The supporting evidence sits in endurance-type, metabolically driven muscle damage, which is what a marathon produces. It does not extend to high-force eccentric loading.

The argument against drinking it every day

A review of nutritional interventions for exercise-induced muscle damage raises a question that applies directly here: "many nutritional interventions described in this review target oxidative stress and inflammation, both thought to contribute to EIMD but are also crucial for the recovery and adaptation process. This calls into question whether long term administration of supplements and functional foods used to target EIMD is indeed best practice." The authors flag "the potential hormetic effect of long term use."

Two things that review does not say. It never names tart cherry — it is a general review of the category, and applying it here is a mechanistic inference, not a finding. And it does not claim that antioxidants ruin training adaptation; it questions whether long-term daily use is best practice.

The practical translation is uncontroversial anyway: use it in race weeks and during dense competition or high-intensity blocks, where fast recovery is the goal, rather than year round while you are trying to build adaptation. If you are unsure which phase you are in, the training load calculator and the recovery planner will place the week for you.

Side effects, and the studies that do not exist

No research has been designed to evaluate the adverse effects of tart cherry juice. That is not the same sentence as "the research found it safe". A search of the literature for adverse events, tolerability or safety reporting turns up no study built to answer the question.

The one quotable safety line comes from a 282-person gout trial: "Adverse events were similar across all three groups." Read the design before leaning on it. The intervention was a tart cherry and citrate mixture rather than plain juice, the trial was open label, every participant was also taking urate-lowering medication, and the comparators were two active alkalinising treatments rather than a placebo. It supports one narrow statement, that this trial saw no between-group difference in adverse events, and nothing wider.

On drug interactions, a search of the human literature returned no interaction studies at all. That cuts both ways: nobody has demonstrated an interaction, and nobody has demonstrated the absence of one. Zero studies is not a clean bill of health.

One gap belongs in a separate category, because it differs in kind. Whether the melatonin in tart cherry interacts with sedative or sleep medication was not searched for this page at all. Its absence here carries no information either way. If you take regular medication, that is a question for a pharmacist or doctor rather than something to infer from a running guide.

Cherries and gout: what is actually known

This is the most searched question about tart cherry and it is not a running question, so here is the state of the evidence without the usual shortcut.

The number everyone repeats comes from a case-crossover study of 633 people with gout, where cherry intake over a 2-day period "was associated with a 35% lower risk of gout attacks compared with no intake (multivariate OR 0.65 [95% CI 0.50-0.85])", and combining cherry intake with allopurinol was associated with a 75% lower risk (OR 0.25 [95% CI 0.15-0.42]). The authors' own verb is "associated with". The design is observational and both the attacks and the intake were self-reported, so this is a correlation and cannot show that cherries prevent anything.

The controlled evidence is thinner and points somewhere less convenient. A crossover trial in 26 overweight and obese adults, who did not have gout, found plasma urate fell from 378.3 to 320.8 micromolar (P < 0.05) on tart cherry juice, while the placebo arm moved from 382.1 to 390.2. And in the only reasonably sized trial in people who actually have gout, 282 men over 12 weeks, urine pH rose and serum urate fell in all three arms with no significant differences between them: the cherry-containing preparation performed like the standard alkalinising treatments it was compared against.

Which leaves the boundary where it has to sit. Gout has diagnostic criteria, a first-line drug and a treatment target: the 2020 American College of Rheumatology guideline recommends allopurinol as the preferred first-line urate-lowering therapy, with treatment titrated to a serum urate below 6 mg/dL. Cherry juice is not that, cannot substitute for it, and delaying proper treatment in its favour would be a poor trade. Whether it earns a place alongside treatment is a question for the doctor managing it.

What about sleep

A systematic review restricted to randomised placebo-controlled trials reports that "the consumption of milk and sour cherries, sources of melatonin, may improve sleep quality in humans." Take the wording literally: the review covers milk and sour cherries together, and its populations were general rather than athletic.

The same review states its own limitation — "lack of homogeneity of treatment dosage and duration" — and closes with "Further studies are necessary." So "may improve sleep quality" is the whole claim. Any specific promise of extra minutes of sleep is going beyond what this evidence supports.

Since sleep is where most of your recovery actually happens, the higher-leverage work is upstream of any drink: see sleep and recovery for runners.

How to actually use it

1. Pick the job first. Post-race strength recovery has the strongest evidence, from a 19-trial pooled analysis. A pre-event dose has thinner support: one crossover trial in cycling plus one meta-analysis pointing the same way.
2. For recovery, run the eight-day protocol: 5 days before the race, race day, and the 48 hours after. The reported amount is about 237 mL of ready-to-drink juice twice a day, which is a secondary-source figure rather than a documented one.
3. For a pre-event dose, time it 90 minutes out. That was the best of the three timings tested (30, 90, 150 minutes), and it matches when the phenolic metabolites peaked in the blood.
4. Expect force, not comfort. Strength returns faster; soreness and creatine kinase do not change in the pooled data.
5. Check what you are buying. Ready-to-drink, dilute-first syrup and drink-neat "concentrate" are three different products at three different doses.
6. Budget the sugar. Roughly 50 g a day for eight days on a standard protocol.
7. Do not drink it year round. Save it for race weeks and dense competition blocks.
8. Clear it first if you take regular medication. The interaction literature here is empty rather than reassuring, and no study has evaluated adverse effects directly.