Ranked by evidence within the category
Physical performance supplements: what the evidence says
This category covers 93 supplements for strength, endurance and training, and the evidence is lopsided: 8 are rated strong, 28 moderate and 57 limited. The strongest cases belong to creatine monohydrate, caffeine and sodium bicarbonate, alongside protein powders such as whey, casein and soy, which work by filling gaps in daily protein intake. Most of the moderate items are variations on those themes, such as buffers, nitrate and citrulline sources, carbohydrate products and multi-ingredient blends. The limited list includes well-known names such as BCAAs, HMB and turkesterone, where the evidence is too thin to rely on.
Evidence last reviewed: September 2026
Strong evidence (8)
One of the most studied supplements in the world.
High-quality, fast-absorbing milk protein.
The world's most widely consumed stimulant.
One of the few sports supplements with a genuine, well-replicated ergogenic effect: it buffers blood acidosis and improves performance in efforts lasting roughly 1-10 minutes.
Casein is a complete, slowly digested milk protein that builds muscle about as well as any other high-quality protein when it fills a gap in daily intake.
Soy is a complete plant protein that produces the same strength and lean-mass gains as whey in head-to-head training trials, and modestly lowers LDL cholesterol.
A filtered whey with roughly 90% protein and very little lactose or fat.
Concentrated carbohydrate in a portable sachet.
Moderate evidence (28)
Raises muscle carnosine, buffering acid during intense exercise.
EAAs are the scientifically defensible version of BCAAs: all nine essential amino acids do drive muscle protein synthesis, and the classic tracer work shows the non-essential ones are unnecessary.
Leucine is the amino acid that switches on the mTOR pathway, and adding it to low-protein meals demonstrably increases muscle protein synthesis.
Citrulline bypasses the gut-and-liver breakdown that cripples oral arginine, making it the more effective nitric oxide precursor.
Beetroot juice raises nitric oxide availability and produces small but real improvements in endurance and repeated high-intensity work, mostly in recreational athletes.
Sodium-containing drinks genuinely improve fluid retention and rehydration after heavy sweat losses, and that part is well established.
Glycerol reliably increases fluid retention and expands body water before exercise, which is a real physiological effect.
Montmorency tart cherry has a legitimate, meta-analysed benefit for recovery of muscle strength after hard exercise — small but real.
Pea protein supports muscle gains when dosed generously - roughly 50 g/day in the one substantial trial - and performs close to whey.
Egg white is a top-tier complete protein (PDCAAS 1.
Carbohydrate during long or hard training genuinely works, but the evidence is for carbohydrate, not for this particular molecule.
Beef protein performs about as well as whey for lean mass and strength when total protein intake is matched.
Multi-ingredient pre-workouts work mostly because of caffeine, and secondarily because of the creatine, beta-alanine and citrulline that some (not all) of them dose properly.
The protein and carbohydrate in a recovery blend genuinely support repair and glycogen resynthesis — but so does food.
Two nitric oxide precursors have real evidence — dietary nitrate (beetroot) and L-citrulline — while L-arginine, the ingredient most "NO booster" labels lead with, is largely destroyed by first-pass metabolism and performs poorly.
Anthocyanin-rich New Zealand blackcurrant extract has one of the better-replicated metabolic effects in sports supplements — it shifts fuel use toward fat during exercise — but translation into actual performance gains is small and inconsistent.
Citrulline bound to malate, the form used in most pre-workouts.
An extracellular buffer, the gentler cousin of sodium bicarbonate.
Whey partially broken into peptides before you drink it, which speeds absorption.
Whole milk protein with the lactose and fat removed, keeping the natural roughly 80:20 casein-to-whey ratio.
Carbohydrate-plus-electrolyte drink mixes for long efforts.
Caffeine delivered through the cheek lining, which gets it into the blood faster than capsules.
Carbohydrate and amino acid drinks sipped during training.
Caffeine-free pump formulas built on citrulline and nitrate sources.
Creatine monohydrate in a chewable format.
A legume protein powder free of soy, dairy and gluten.
Pre-digested amino acids covering the full spectrum.
A Thai rhizome rich in methoxyflavones, sold for stamina and fat loss.
Limited evidence (57)
BCAAs are one of the best-selling and worst-supported supplements in sports nutrition: leucine, isoleucine and valine alone cannot build muscle because the other six essential amino acids are missing.
HMB is a leucine metabolite with an unusually polarised literature: early industry-linked trials reported large gains that independent meta-analyses in trained young people have failed to reproduce.
Ketone salts do raise blood beta-hydroxybutyrate, but far less than ketone esters, and the performance literature is overwhelmingly null or negative.
Betaine reliably lowers homocysteine and is an approved drug for homocystinuria, but its popular use as a strength and power supplement rests on small, conflicting trials.
ZMA is zinc, magnesium and B6 in a branded ratio, sold on the promise of higher testosterone and better sleep.
Cordyceps has a handful of small human trials suggesting modest gains in ventilatory threshold and high-intensity exercise tolerance, but nothing approaching a convincing evidence base.
Ecdysterone's reputation rests almost entirely on a single small German trial that reported surprisingly large lean-mass and strength gains.
There is no published randomised controlled trial of turkesterone in humans for muscle, strength or any other outcome; the entire case is rodent data plus extrapolation from its chemical cousin ecdysterone.
A caffeine-like purine alkaloid pitched as "caffeine without the tolerance.
A single small 8-week trial suggests a large 48 g dose of rice protein isolate matches whey for body composition and strength - and that trial is essentially the whole direct evidence base.
Hemp protein is a fiber-rich but moderate-quality plant protein with measured protein quality well below whey, casein or soy, and no human trials at all on muscle, strength or body composition.
Dehydrated coconut water is a decent potassium source and a palatable rehydration drink, but head-to-head trials show it performs about the same as water or a sports drink — and its sodium content is too low for serious sweat replacement.
AAKG is L-arginine bonded to alpha-ketoglutarate, sold on the promise of a bigger nitric oxide "pump" than plain arginine.
Creatine hydrochloride dissolves far better than monohydrate, which is a real chemistry advantage and a real convenience — but it does not translate into better muscle creatine loading, more strength, or lower dosing.
Kre-Alkalyn is pH-buffered creatine sold on the premise that monohydrate degrades into useless creatinine in the stomach — a premise that is false.
Creatine ethyl ester was designed for better membrane permeability, but chemistry and clinical data both went against it: at stomach pH it degrades rapidly into inert creatinine, and the one controlled training trial found it raised muscle creatine less than monohydrate while dramatically raising serum creatinine.
Silk amino acids are hydrolyzed silk fibroin protein, marketed for stamina, muscle growth and libido.
Oral ATP does not survive digestion intact, and the impressive strength and hypertrophy results come almost entirely from one industry-linked research network with no independent replication.
Controlled trials in athletes found no gains in aerobic power, strength or hormones beyond training alone.
A cocoa flavanol sold as a natural myostatin inhibitor.
An omega-6 fatty acid sold to amplify the inflammatory signal that drives training adaptation.
A soy-derived lipid that activates mTOR in cell culture.
Sold as a natural plant anabolic, the compound in these products is a synthetic steroid with zero human trials.
A mass gainer is maltodextrin, sugar and whey sold at a premium as a muscle-building product.
Grass-fed whey is ordinary whey protein from pasture-raised cows.
Sold as a 'Russian secret' anabolic because it contains ecdysteroids, but a typical 500 mg capsule delivers only a few milligrams of them — far below the doses used in the one human ecdysterone trial.
A Siberian ecdysterone source with a Soviet sports-science legend behind it and almost no modern human data.
D-ribose is sold on a plausible mechanism — refilling depleted adenine nucleotide pools — but controlled trials in trained people consistently find no performance benefit, and the fatigue and fibromyalgia data everyone quotes come from an uncontrolled open-label study.
Inosine was a 1980s-90s sports supplement that failed every controlled performance test, then had a second life as a urate-raising neuroprotectant for Parkinson's disease — until a phase 3 trial was stopped for futility.
Pangamic acid is not a vitamin, has no agreed chemical identity, and failed the exercise trials that were run in the early 1980s.
One of the three branched-chain amino acids, sold alone for glucose uptake.
The third branched-chain amino acid, marketed for endurance and recovery.
A rapidly absorbed carnitine salt used for recovery rather than fat loss.
A non-proteinogenic amino acid added to pre-workouts to inhibit arginase and raise nitric oxide.
Creatine bound to magnesium, sold as an upgraded form.
Creatine paired with nitrate for solubility and a nitric-oxide angle.
A bonded arginine-silicate complex that raises blood arginine more than arginine alone.
A high-molecular-weight starch marketed as a fast carbohydrate for refuelling.
A slowly digested sugar sold as steady endurance fuel.
Phosphate loading before competition, an old idea with a stubbornly inconsistent literature.
The body's direct precursor to creatine.
A minor leucine metabolite sold for lean mass and reduced soreness.
The keto acid of leucine, marketed for recovery and anti-catabolic effects.
A plant protein made from defatted pumpkin seed, with a decent amino acid profile but low lysine.
Powdered whole cricket or cricket protein isolate, a complete protein with an environmental argument behind it.
Drinks that raise blood ketones without a ketogenic diet, sold as an alternative fuel for endurance.
A purine alkaloid related to caffeine and theacrine, sold in pre-workouts for energy and focus.
A triterpenoid from apple peel that drew attention after rodent studies showed muscle growth and fat loss.
A methyl donor from the choline-betaine pathway, promoted since the 1980s for endurance and immunity.
A Krebs cycle intermediate found in apples, used for fibromyalgia pain and as the malate half of citrulline malate.
Sodium tablets for long, hot endurance events.
Whey from goat milk, marketed as a gentler dairy alternative.
An amino acid salt used in hospital nutrition for burns and wounds.
A hypoallergenic seed protein used in allergen-free powders.
Magnesium and electrolyte blends sold for night cramps.
Rhodiola and antioxidant blends sold for acute mountain sickness.
A long-chain alcohol from wheat germ and sugar cane wax, sold for stamina and endurance since the 1970s.
Comparison guides for this category
Whey, casein, soy, pea, rice, hemp or egg white?
Monohydrate, HCl, Kre-Alkalyn or ethyl ester?
L-carnitine, acetyl-L-carnitine, L-carnitine L-tartrate or propionyl-L-carnitine?
Concentrate, isolate, hydrolysate, grass-fed or goat whey?
Plain caffeine, caffeine gum or caffeine + L-theanine?
Are either worth taking on a normal diet?
L-arginine, arginine silicate or AAKG?
Plain L-ornithine or ornithine alpha-ketoglutarate?
Pure citrulline or the malate blend in most pre-workouts?
Trimethylglycine for homocysteine, or betaine HCl for stomach acid?
Creatine, beta-alanine or HMB?
BCAAs, EAAs or L-glutamine for muscle?
Electrolyte powder or coconut water powder for rehydration?
Goals these supplements are ranked for
An evidence-ranked guide to supplements for muscle and strength: what actually works, what works only in specific cases, and what to skip.
An evidence-ranked guide to endurance supplements: carbohydrate and caffeine first, what helps only in specific events, and what to skip.
An evidence-ranked guide to recovery supplements: what speeds the return of strength, what eases soreness a little, and what does nothing.
An evidence-ranked guide to hydration and electrolyte supplements: what helps after heavy sweating, what is situational, and what to skip.
An evidence-ranked guide to supplements for older adults: what actually works, what helps only in specific cases, and what to skip.
What actually helps with low energy and fatigue, ranked by strength of evidence, plus the popular energy supplements that do not hold up.
Our verdict
Start with creatine monohydrate, enough total protein and, if you tolerate it, caffeine; for hard efforts lasting a few minutes, beta-alanine or sodium bicarbonate are the next step, though bicarbonate can upset the gut. Be sceptical of pre-workouts and recovery blends that charge for a long ingredient list: their effect usually comes from two or three proven parts, and proprietary blends often under-dose them.