Methodology

Carbohydrate, sodium, calories and every other figure is read directly off the product's own label — the same panel you'd read on the packet. We don't estimate a number because a competitor discloses it and this product doesn't.

The one figure that sometimes needs interpretation is the glucose:fructose ratio, since not every brand states it outright. Each product's ratio is tagged with where it actually came from:

• Stated — the brand publishes the ratio directly.
• Derived — computed from ingredient percentages the brand does disclose.
• Estimated — inferred from what's on the panel when neither of the above is available. It's marked as an estimate everywhere it appears, but it scores the same as a Stated ratio — there's no confidence discount built into the number itself.
• Not disclosed — the brand doesn't say, and we don't guess.

Every score on this site is absolute, not relative — each dimension (absorption rate, carb density, sodium, cost per gram, gut comfort for gels, potassium/calcium/magnesium for electrolytes) is scored against a fixed scale, not against whatever else happens to be in the catalog. A 90 means the same thing today as it will after the next product is added; scores only change when a product's own declared data changes.

Where real sports-nutrition reference points exist, the fixed scale is anchored to them: carb density tops out at 90 g, matching the upper end of the range cited for multi-transportable-carb intake; sodium scales top out at 1000 mg, the upper end of what's typically lost to sweat in an hour; potassium and magnesium follow the same logic at smaller scales. Hover any dimension label on a product page for its specific scale and reasoning.

Some dimensions don't have a physiological anchor to reach for and say so plainly: both cost dimensions are scored against a market-based price range, not a nutrition benchmark, and calcium's scale is anchored to this catalog's practical ceiling rather than sweat-loss research, since calcium loss during exercise is minor and isn't really a target most products are optimizing for.

A product's overall score is the plain average of whichever dimensions it has real data for. Missing data is left out of the average, not counted against it — a product that doesn't disclose sodium isn't penalized for the gap, it's just scored on what it does disclose. Whenever that's happened, the product page says so directly: "Based on N of X measures" next to the score.

Gut comfort (gels only) is a proxy, not a taste or tolerance test: it's scored from ingredient count on a fixed 1-to-15 scale, on the reasoning that a shorter, simpler label is generally easier on the stomach. It's disclosed as what it is — an inference from the label, same as everything else here — not a stand-in for someone actually racing on it.

Prices are tracked from a specific retailer, linked directly on every product page, and reflect what that retailer listed as of the review date shown at the top of the page. Prices move; we don't re-check every listing daily, so treat the number as a recent snapshot and the link as the source of truth.

None of the links on this site are affiliate links. We don't earn anything when you click through or buy, and that's on purpose — it removes any incentive to rank a product higher because it pays better.

Product photos are sourced only from the official brand site or an authorized retailer — never stock photography, never a competitor's marketing image. Where a brand's own photography wasn't clean enough to use as-is, we've edited the background, never the product itself.

The date on each product page reflects when that product's written review — its verdict, pros, cons and data — was last actually verified or revised, not when the page was last deployed. A photo swap or a copy-editing pass elsewhere on the site doesn't move a product's review date; a change to what the review actually claims does.

22 gels, 14 drink mixes and 11 electrolyte products, and growing. Don't see something you use? Search for it — if it's not here yet, that tells us it should be.