What it's actually good for
Raising your blood ketones. That part works, it works within an hour, and it is the only claim on this page that survives contact with the evidence.
Start with what is in the bottle, because most reviews get this wrong. Ketone-IQ sells 10 g of chirally pure (R)-1,3-butanediol per serving. No ester, no BHB salt, no MCT oil. The company sold a ketone ester first, under the H.V.M.N. name, and moved off it; the brand's own changelog cites the harsh taste and the acid load of the ester bond as the reason. Your liver converts the diol to beta-hydroxybutyrate using alcohol dehydrogenase. The conversion is real and it is slower and lower-peaking than swallowing an ester of equivalent mass.
That matters because almost every trial you will see cited on a ketone product page used the ester.
What the research says
Endurance performance (Grade C). The 2022 meta-analysis by Brooks and colleagues in the International Journal of Sport Nutrition and Exercise Metabolism is the number to know. They screened 569 papers, pooled the 8 randomized trials that qualified, and across 80 participants found a Hedges g of 0.136 (95% CI -0.195 to 0.467, p = .419). Zero. Nothing in the time-to-exhaustion sub-analysis, nothing in the time-trial sub-analysis.
Set against that, the famous positive study. Cox et al. published five sub-studies of 39 high-performance athletes in Cell Metabolism in 2016, and the fuel-metabolism findings are solid work: less muscle glycolysis, lower lactate, more intramuscular fat oxidation. But the performance headline came from study 5 only, which enrolled 8 athletes. They cycled 411 ± 162 m further over a 30-minute time trial, a 2% mean gain. Eight athletes, one lab, one molecule that is not the one sold here. That is the entire empirical foundation of "ketones make you faster."
Two trials point the other way. Leckey and colleagues gave 10 professional cyclists a ketone diester before a simulated World Championships time trial in 2017 and they finished 2 ± 1% slower, with gut discomfort and higher perceived effort. In 2025, Stalmans and colleagues put 13 men at 4,000 m simulated altitude and found ketone ester cut peak power by 3.6%.
Focus and cognition (Grade C). This is now the main pitch, and it is the thinnest claim here. Nobody has run a placebo-controlled cognitive trial on Ketone-IQ or on (R)-1,3-butanediol by itself. The closest data is the 2025 altitude study above, where the ketone ester failed to protect cognition in conditions designed to degrade it. Ketone-IQ's marketing leans on a US Special Operations Command research contract. A defence contract is a procurement fact, not a result, and it should not be read as evidence.
Appetite and weight (Grade C). Here the literature is unusually clean about its own limits. Stubbs et al. 2018 gave 15 adults a ketone ester and saw ghrelin, GLP-1 and PYY drop for 2-4 hours, with self-reported hunger down at 90 minutes. Okada et al. then measured the thing that matters: 13 adults, ketone ester after exercise, ghrelin down again (p = 0.002), and ad libitum energy intake over the next two hours unchanged (p = 0.488). Hormones moved, eating didn't. No trial has weighed anyone.
Recovery from training overload (Grade C). The one result worth watching. Poffé's group ran athletes through 3 weeks of deliberate overload with a ketone ester and blunted the overreaching symptoms the control arm developed. It drew a published critique and a reply in the same journal, it used the ester, and no independent lab has repeated it. Promising; not settled.
How much, and which form
Ketone-IQ's serving is 10 g of the diol, taken 10-30 minutes before effort. Cox's positive sub-study used 573 mg/kg of monoester, about 43 g for an 80 kg rider, and the cyclist trial used 2 × 250 mg/kg of diester. Those numbers do not convert. A gram of diol is not a gram of ester, and the whole category trades on readers assuming it is.
If you are buying on the endurance research, note that you cannot buy what the research used at a normal price. If you are buying on the biomarker, 10 g will move it.
Safety & interactions
GI distress is the main adverse effect and it scales with dose. That is not a minor tolerability note: in the Leckey cyclist trial, gut discomfort was the likely mechanism behind the slower times, so the side effect and the null result are the same event. The 2023 open-label trial of (R)-1,3-butanediol found 34.5 g across an hour produced no or mild GI symptoms in most of 26 adults, no increase in sleepiness on the Stanford scale, and mild headache in five. Blinding is a real problem across this literature; the taste is distinctive enough that participants often know which arm they are in.
Anyone with type 1 diabetes should talk to a clinician first, since exogenous ketones raise blood ketones regardless of insulin status and can muddy the readings used to catch ketoacidosis. The diol is cleared by alcohol dehydrogenase, so stacking it with alcohol puts two substrates on one enzyme, untested. No pregnancy, nursing, or paediatric data. No long-term data on this molecule at all. This is informational only, not medical advice.
How we picked the brand
Nothing in this category has demonstrated an outcome, so the bar is label honesty and manufacturing hygiene rather than results: the active compound and its dose stated plainly with no proprietary blend, NSF Certified for Sport with per-batch third-party testing available on request, and a listing sold by the manufacturer rather than a reseller. Ketone-IQ clears that bar and is the most transparent product in a category where most competitors will not tell you whether you are buying an ester, a diol, or a salt.
We are linking it because people are going to buy it. What you are paying roughly $3 a serving for is a reliable rise in blood BHB. If you want a performance supplement with an outcome behind it rather than a biomarker, creatine monohydrate costs about a tenth as much and carries a Grade A on this site.