Sodium Benzoate, Sorbic Acid, and TBHQ: What Ingredients Are Preservatives, Ranked by Evidence

Why Does the Evidence Ranking of Preservatives Matter?

When you ask what ingredients in a product are doing the preservation work, the name on the label tells you very little about how well that substance has been tested on real humans over real time spans. Regulatory approval and genuine scientific confidence are not the same thing.

An ingredient can be on the FDA's Generally Recognized as Safe list, or hold an E-number under EFSA review, and still have most of its safety record built on animal feeding studies from the 1970s. The ranking here is about the quality and depth of that record, not about whether an ingredient is currently permitted.

Understanding this distinction helps you decide how much weight to give a label. It also gives context to the ongoing re-evaluation work EFSA has been publishing through the mid-2020s on legacy additives.

Sorbic Acid and Potassium Sorbate: The Gold-Standard Evidence Tier

Sorbic acid, and its salt potassium sorbate, sit at the top of the evidence ladder. Human dietary exposure has been measured in multiple large national nutrition surveys, and the Joint FAO/WHO Expert Committee on Food Additives has reviewed the compound in several full re-evaluations. EFSA completed a detailed opinion in the early 2010s and updated exposure estimates using consumption data from more than a dozen European countries.

The mechanism is well understood: sorbic acid disrupts the cell membranes of yeasts and moulds at the concentrations used in bread, cheese, and juice. At those same concentrations, the human body metabolises it through the same beta-oxidation pathway it uses for naturally occurring fatty acids, so there is no exotic metabolic burden.

No credible regulatory body has flagged a concern at current approved use levels. That does not mean the science is complete — some researchers have noted that interaction effects with other additives in the same product are understudied — but as single-ingredient files go, this one is unusually thorough.

Sodium Benzoate: Well-Studied but With One Open Question

Sodium benzoate has been used as a food preservative for well over a century, which means its basic toxicology file is extensive. Regulatory agencies in the US, EU, and UK have all reviewed it multiple times, and acceptable daily intake figures are established by both the FDA and EFSA.

The open question involves benzene. When sodium benzoate is present in the same acidic drink as ascorbic acid (vitamin C) and exposed to heat or light, a small amount of benzene can form. The FDA has monitored this since the mid-2000s, publishing survey data on soft drink samples. Most products tested below the drinking-water action level, but the reaction pathway is real and not yet fully characterised across all product formats.

This does not make sodium benzoate uniquely dangerous — the exposure in a typical serving is very small — but it is a genuine area where the science remains active rather than settled. Any ingredient analyzer worth using should flag this nuance rather than giving a flat pass or fail.

TBHQ: A Solid Dossier, Thinner Independent Record

Tertiary butylhydroquinone, almost always listed as TBHQ, is an antioxidant preservative used mainly in oils, crackers, and fast-food frying fats. It prevents rancidity by interrupting the lipid oxidation chain. The FDA permits it at no more than 0.02 percent of the fat or oil content in a food, and EFSA has also set an acceptable daily intake.

The safety dossier behind that approval is real, but a notable share of the foundational studies were conducted or funded by the food industry rather than by independent academic groups. EFSA's 2020 re-evaluation acknowledged this, noting that the committee relied heavily on data submitted through regulatory channels rather than peer-reviewed literature. The agency concluded the current intake was safe but called for updated independent studies.

TBHQ therefore sits in a middle tier: regulated with numbers, mechanism understood, but with a thinner body of independent human-exposure work than sorbic acid or sodium benzoate.

Natamycin: Effective, Permitted, but Narrowly Studied

Natamycin is a naturally produced antifungal compound, made by a soil bacterium, and used as a surface treatment on hard cheese, dried sausage, and some baked goods. Because it is produced biologically, it sometimes escapes scrutiny on ingredient checkers that flag only synthetic additives — but it is a preservative in every functional sense.

The human data for natamycin is narrower than for the top-tier preservatives. Most published safety reviews focus on its antimicrobial pharmacology and on the theoretical concern that dietary exposure could contribute to antifungal resistance in clinical settings. EFSA reviewed natamycin as E235 and found it acceptable at current use levels, but the exposure modelling relied on a smaller pool of consumption studies than is typical for older additives.

The antifungal-resistance question is genuinely unresolved. Researchers publishing in food microbiology journals through the early 2020s noted that the topic had not been studied with the kind of controlled longitudinal design that would settle it. That gap keeps natamycin out of the top evidence tier.

Rosemary Extract and Mixed Tocopherols: 'Natural' Does Not Mean Well-Tested

Rosemary extract and mixed tocopherols (a form of vitamin E) are the preservatives most likely to appear on a product marketed as clean-label or free from artificial additives. They are real antioxidants that slow fat oxidation, and they are permitted globally.

The challenge is that the safety data for these ingredients in their preservative function — meaning the specific preparations, concentrations, and food matrices used industrially — is much thinner than for synthetic counterparts. Tocopherols as nutrients have an enormous literature, but the question of is this ingredient safe as a preservative at the concentrations used in a shelf-stable product is less thoroughly answered.

Rosemary extract in particular contains compounds such as carnosic acid and rosmarinic acid that have shown hormetic effects in cell-culture studies — beneficial at low doses, potentially disruptive at high doses — but human dietary exposure data at real food-use levels is limited. The 'natural' label carries genuine meaning in terms of origin, but it does not substitute for a full regulatory evidence file.

How Do the Preservatives Compare at a Glance?

The table below summarises each preservative's evidence tier, the type of food it typically appears in, and the main open question regulators or researchers have identified.

Using a food ingredient scanner like InZoRAH against a label can surface which of these preservatives are present, but the tier rating here reminds you that finding a preservative is just the start of the question — the depth of its safety record matters too.

What Should You Actually Do With This Information?

No ingredient on this list is classified as unsafe by any major food authority as of November 2026. The ranking is about confidence and completeness of evidence, not about immediate harm. For most people eating a varied diet, the exposure from any single preservative in a single product is small.

Where the ranking becomes practically useful is in comparing two similar products. If one uses sorbic acid and the other uses rosemary extract, the sorbic acid product actually has the deeper human-safety record — which is the opposite of what most shoppers assume. That is what ingredients research often reveals: the conventional choice is sometimes the better-understood one.

Always read the physical label if you have a known allergy or sensitivity, since formulations change. InZoRAH pulls from Open Food Facts and openFDA enforcement records to flag recent reformulations, but no app replaces reading the pack in hand.