We expected to grade this one B.
The Sleepy Girl Mocktail — tart cherry juice, a scoop of magnesium powder, topped with sparkling water — is the rare viral wellness trend with plausible mechanisms behind both active ingredients. It is cheap, it is not dangerous, and unlike most of what we cover, nobody is getting hurt.
Then we read the trials, and B stopped being defensible.
The tart cherry evidence is two tiny pilots
The headline figure is that tart cherry juice adds 84 minutes of sleep. That comes from a 2018 crossover trial that randomised 11 people, excluded three for sleep apnoea, and analysed eight. Total sleep time was the only one of six polysomnography measures to reach significance — sleep efficiency, sleep latency, wake after sleep onset, number of awakenings and REM latency were all non-significant. The standard deviation on that 84 minutes was ±62.
The authors were explicit that it was "a pilot study to determine power for a larger, adequately powered study." It was supported by the Cherry Marketing Institute, which also supplied the juice. It has not been replicated.
The other insomnia trial, from 2010, had 15 participants, used sleep diaries rather than polysomnography, and found only wake-after-sleep-onset significantly reduced — not total sleep time, and not sleep latency. A frequently repeated "17 more minutes of sleep" from that study is a misreading: 17 minutes less time awake, which is not the same thing.
The one systematic review, published in 2025, declined to run a meta-analysis at all and concluded the evidence "is still limited and heterogeneous."
The melatonin argument is off by orders of magnitude
Tart cherries do contain melatonin. Montmorency cherries measure about 13.5 nanograms per gram — so 100 g supplies roughly 1.35 micrograms.
A melatonin supplement is typically 0.5 to 5 milligrams. The cherry juice is delivering something in the range of one-thousandth of a supplement dose. Whatever tart cherry does, "it's the melatonin" is not a serious explanation.
The magnesium evidence is thinner than its reputation
The meta-analysis everyone cites pooled three trials in 151 older adults and found sleep onset latency reduced by about 17 minutes. Total sleep time increased by 16 minutes — not statistically significant. The authors' own assessment: all trials at "moderate-to-high risk of bias," outcomes supported by "low to very low quality of evidence."
Two details rarely survive the retelling. Those trials used magnesium oxide and citrate — not the glycinate the recipe calls for. And the trial driving the latency result required participants to have low or low-normal serum magnesium to enrol. It is evidence about correcting a shortfall, not about topping up someone already replete.
The only glycinate-specific sleep trial reported a p-value of 0.049 and a small effect size, with every secondary outcome null and no objective sleep measurement — and the placebo group improved by 18% on its own.
Nobody has studied the drink
This is the part that settles the grade. No trial has tested the combination. Any claim of synergy between tart cherry and magnesium is unevidenced.
Worse, the recipe doesn't match the research it borrows credibility from. The trials used 240 ml of tart cherry juice twice daily; the viral recipe uses half a cup once at bedtime — roughly a quarter of the studied dose. The suggested 100–200 mg of magnesium is below every trial dose.
Two practical notes
A full serving of some magnesium powders delivers 350 mg of elemental magnesium — exactly the NIH's upper limit for supplemental magnesium, before counting food. That limit exists because of gastrointestinal effects, and the risk of toxicity rises with impaired kidney function. Anyone with reduced renal function should not be free-pouring magnesium.
Magnesium also interacts with medications: separate it from bisphosphonates and from tetracycline or quinolone antibiotics by several hours.
So what is it?
Probably a pleasant bedtime ritual that helps some people relax, with a real possibility that the ritual and the expectation are doing most of the work. A sleep physician quoted on the trend put it plainly: "It is possible that this mocktail works as a placebo. The placebo effect is quite strong."
That is not a takedown. If a nightly wind-down routine helps you sleep, the mechanism is your business. But "there's science behind it" overstates what has actually been shown, and the honest grade for a claim with plausible mechanisms, tiny industry-linked pilots, under-dosed recipes and no test of the actual product is Mixed / Unproven — not Mostly True.
We would happily revise this to a B if someone runs the trial. One properly powered tart cherry insomnia study is currently underway, which itself tells you what the field thinks of the existing evidence.