It depends entirely on which “real recipe” you mean — and the two answers could not be further apart.
The 1886 Pemberton original
Is a genuine early formula already in the public record?
โlikely — range 75–90%
The Coke in a can today
Does any published recipe match the current commercial product?
โeffectively no — range 1–5%
A real early Coca-Cola formula is, for practical purposes, already public. The real current formula is not — and even a perfect copy of it could not be legally reproduced by an outsider. Everything below is why.
01 / PremiseWhy “the real recipe” is a moving target
There is no single fixed object called “the Coca-Cola formula.” The recipe has been materially changed at least four documented times, and today it exists as several products at once. Any leaked recipe can, at best, match one moment or one region.
Candler’s 1891 reformulation. Asa Candler altered Pemberton’s list — “by most accounts improving the flavor” — and built the secrecy regime. The formula had already drifted before any famous leak was recorded.
Cocaine removal, ~1903–1929. Not one event. Fresh coca leaf was dropped around 1903 for a “spent” decocainized extract; trace cocaine lingered until roughly 1929. The coca component changed form — it never left.
Sugar → HFCS, 1980–1984. US Coke moved from cane/beet sugar to high-fructose corn syrup (up to 50% allowed Jan 1980, 100% by Nov 1984). Today’s US product differs from every pre-1980 recipe in its primary sweetener.
New Coke, 1985. Proof the company reformulates at will — launched in April, reverted in ~78 days, and the “Classic” relaunch kept HFCS rather than restoring cane sugar. Even “Classic” is a changed product.
Many variants, right now. Mexican Coke and Passover “yellow-cap” Coke use cane sucrose; caffeine-free omits caffeine; bottlers worldwide tune water, sweetener, and carbonation to local law. “The current product” is itself plural.
02 / The leakWhat the famous recipe actually is
The 2011 “leak” by the radio program This American Life (Episode 427) is not one discovery but one formula surfacing through two independent documents that agree almost exactly — the signature of a genuine common ancestor, not a hoax:
A handwritten page in Everett Beal’s recipe book, photographed in a February 1979Atlanta Journal-Constitution column.
A page from John Pemberton’s own papers, published by historian Mark Pendergrast in For God, Country and Coca-Cola (1993).
The two blends are identical except that Pendergrast’s is quoted at a clean 4× scale. Here is the Beal version, verified against the official This American Life PDF:
℞ Coca-Cola — Beal recipe bookper batch
Main syrup
Fluid extract of coca (USP)
3 drams
Citric acid
3 oz
Caffeine
1 oz
Sugar
30 lbs
Water
2.5 gal
Lime juice
2 pt 1 qt
Vanilla
1 oz
Caramel (for color)
1.5 oz+
“Merchandise 7X” flavor — 2 oz per 5 gal of syrup
Alcohol
8 oz
Orange oil
20 drops
Lemon oil
30 drops
Nutmeg oil
10 drops
Coriander oil
5 drops
Neroli oil
10 drops
Cinnamon oil
10 drops
Note from the TAL document: “The Pemberton formula for 7X is the same as the Beal, just four times as much.” The open-source colas (OpenCola, 2001) descend from this same oil lineage but make no claim to be Coke.
The show itself never claimed this was the modern drink — only Pemberton’s original or a close early version. Coca-Cola’s own archivist, Phil Mooney, went on record:
“Could it be a precursor? Yeah, absolutely. Is this the one that went to market? I don’t think so.”Phil Mooney · Coca-Cola archivist
03 / The wallWhy a paper recipe is not the product
Even a perfectly transcribed, perfectly authentic formula would not let an outsider reproduce Coca-Cola. The recipe under-determines the product in three independent ways:
An ingredient you cannot legally buy. Coke still uses a decocainized coca-leaf extract (“Merchandise No. 5”). Exactly one company — Stepan, in Maywood, New Jersey — is DEA-licensed to import coca leaf (~100 metric tons/yr from Peru), and it sells the extract only to Coca-Cola. A competitor cannot obtain it at all.
A base chemistry that reshapes the flavor. At Coke’s ~2.5 pH, phosphoric acid (not the leaked citric) changes how the same oils register; the proprietary caramel and the sugar’s viscosity are load-bearing, not cosmetic.
A process that isn’t on the page. Commercial syrup is chemically aged — hydrolysis and isomerization round off harsh raw terpenes — then forced into sub-micron weighted emulsions by high-pressure homogenizers. GC-MS reveals the makeup, never the method.
04 / The proofWhat happened when people actually made it
The insider taste test failed.This American Life had the leaked recipe manufactured; tasting it, Mooney said it “doesn’t have what we call the bite and the burn.” The first batch tasted like “orange-flavored baby aspirin.”
“Nobody could tell the difference” is a half-truth. In the blind supermarket test, 28 of ~30 shoppers (93%) correctly identified the real Coca-Cola. Many simply preferred the recreation — preference is not identity.
Home cooks land “cola-adjacent.” A food historian who brewed the 1886 formula rated it 7.1/10, “in the Coca-Cola family, more complex and more herbal,” with a floral neroli note absent from real Coke.
The one convincing clone used no paper recipe. The 2026 LabCoatz recreation reverse-engineered the finished drink by GC-MS — finding compounds the leaked recipes never mention — which is itself evidence the paper formulas are not the modern product. (Self-reported, no independent flavor-chemist verification.)
05 / ScorecardMyth versus documented fact
Famous claim
Verdict
Basis
Only two people know it; they never fly together; each knows half
Myth
Company says “a few people”; no travel policy exists; Snopes debunks the half-each version.
The written formula sits in a vault
Fact
Guaranty Trust NY (1919) → Trust Company / SunTrust Atlanta (1925) → public display vault since Dec 8, 2011.
The 2011 recipe is the Coke sold today
Myth
Even This American Life disclaimed this; a precursor at best.
The 2011 recipe is a genuine early Pemberton formula
Likely
Two independent provenance chains; archivist granted “could be a precursor.”
Oils consistent across sources; exact proportions authenticated by no one.
Coca-Cola left India (1977) rather than disclose the formula
Fact
FERA demanded equity dilution + disclosure; Coke exited for ~16 years.
A court once ordered Coke to reveal the formula
Partly
1980s Delaware discovery order — under seal, never public; case settled.
The secret recipe is the company’s real moat
Myth
The enforceable value is the perpetual trademark + the coca chokepoint + bottling scale.
Final verdict
The probability that a known, published recipe is the actual current commercial formula is very low — on the order of 1–5%, and the claim is close to ill-formed, because the current product is itself plural and has changed repeatedly since any leaked recipe was written.
The probability that a genuine early Coca-Cola formula (Pemberton-era, ~1886–1891) is already public is high — roughly 75–90%. It is not the drink in the can today, but as a historical artifact it is very probably real.
So: yes to the 1886 ancestor, no to the modern product. The enduring “secret” is less a locked chemical mystery than a trademark the law protects forever, a single legally throttled ingredient no competitor can source, and a brilliantly maintained piece of marketing theater layered on top of a trade secret that is genuine — but far less unknowable than the legend insists.
SourcesSelected references
This American Life, Episode 427 “Original Recipe” — transcript & recipe PDF. thisamericanlife.org/427
Mark Pendergrast, For God, Country and Coca-Cola (1993).
NBC News, “Coca-Cola says found recipe isn’t the real thing.” nbcnews.com
NPR, “The Coca-Cola Recipe Revealed By This American Life.” npr.org