Is dark chocolate actually good for you?

One narrow effect is authorised in the EU: cocoa flavanols help maintain the elasticity of blood vessels, contributing to normal blood flow, at 200 mg of cocoa flavanols daily. That is a real, substantiated physiological effect. It is not the same as a finding that dark chocolate prevents disease, lowers blood pressure as a treatment or extends life, none of which this authorisation covers.

How much dark chocolate would I need to eat?

The EFSA opinion described the 200 mg daily condition as obtainable from about 10 g of high-flavanol dark chocolate or about 2.5 g of high-flavanol cocoa powder. The catch is the words high-flavanol: unless a product declares a measured flavanol content, there is no way to know whether it reaches that level, and most bars do not declare one.

Does a higher cacao percentage mean more flavanols?

Not reliably. Percentage describes the share of cacao-derived material, while flavanol content depends heavily on processing — fermentation, roasting and especially alkalization all reduce it. An 85 per cent bar made from heavily processed cocoa can carry fewer flavanols than a lower-percentage product made to retain them.

Is Dutch-process cocoa lower in flavanols?

Alkalization is the processing step most associated with flavanol reduction, because the treatment that softens acidity and darkens colour also degrades these compounds. If flavanol content is the reason you are buying cocoa powder, natural rather than alkalized is the more consistent choice, though only a declared measured figure settles it.

Evidence / Nutrition claims · 9 minute read

Cocoa flavanols:
one narrow authorised claim.

Dark chocolate is widely described as healthy. Exactly one cocoa claim is authorised in the EU, it is far narrower than the headlines, and most chocolate does not meet its conditions.

Flavanols originate in the seed and are substantially reduced by ordinary processing. Editorial reconstruction, not analytical evidence.

The short answer

Know this first.

There is one authorised EU health claim for cocoa flavanols: that they help maintain the elasticity of blood vessels, which contributes to normal blood flow. It rests on a demonstrated effect on endothelium-dependent vasodilation and requires 200 mg of cocoa flavanols daily, obtainable from about 2.5 g of high-flavanol cocoa powder or 10 g of high-flavanol dark chocolate. It is not a claim about heart disease, blood pressure, weight, mood or longevity, and an ordinary supermarket dark bar is not automatically a high-flavanol product.

One effect, one dose, one wording.

What was actually authorised

In 2013 the European Commission authorised a health claim for cocoa flavanols following a scientific opinion from the European Food Safety Authority. The authorised wording is that cocoa flavanols help maintain the elasticity of blood vessels, which contributes to normal blood flow. The condition of use is 200 mg of cocoa flavanols consumed daily, which the opinion described as obtainable from roughly 2.5 g of high-flavanol cocoa powder or 10 g of high-flavanol dark chocolate.

The mechanism assessed was endothelium-dependent vasodilation — the ability of the vessel lining to signal the vessel to widen. That is a real, measurable physiological endpoint, and the assessment concluded a cause-and-effect relationship had been established for it in the general population. It is also a much narrower endpoint than the phrase good for your heart implies, and the distance between the two is where most chocolate health writing lives.

Authorised in 2013 following an EFSA scientific opinion.
Condition of use: 200 mg of cocoa flavanols per day.
The assessed endpoint is endothelium-dependent vasodilation, nothing broader.

High-flavanol is a specification, not a synonym for dark.

Why your bar probably does not qualify

The claim is conditioned on high-flavanol cocoa, and that is a compositional specification rather than a description of any dark chocolate. Flavanols are bitter and astringent, and much of conventional processing has historically been aimed at reducing exactly those qualities. Fermentation, drying, roasting and above all alkalization — the Dutching step that softens acidity and darkens colour — all reduce flavanol content, sometimes substantially.

So a bar can be 85 per cent cacao, taste intensely bitter, and still carry a modest flavanol load, while a product engineered and analysed for flavanol retention carries a much higher one. Unless a product declares a measured flavanol content, you cannot infer it from the percentage, the colour, the price or the word dark. Most bars do not declare it, because most bars were not made to.

Alkalized cocoa is systematically lower in flavanols.
Cacao percentage is not a proxy for flavanol content.
Without a declared, measured figure, the 200 mg condition cannot be verified by the buyer.

The gap between an authorised claim and a headline.

What the claim does not say

An authorised claim is a narrow permission to state one substantiated effect. It is not a finding that chocolate prevents heart attacks, lowers blood pressure as a treatment, aids weight loss, improves mood or extends life. Those are separate questions with their own, generally weaker or more contested, evidence bases, and none of them was authorised here.

It also says nothing about the rest of the food. A 10 g portion of high-flavanol dark chocolate arrives with sugar, fat and energy, and a claim about one compound does not settle whether the product is a good idea in any particular diet. The honest framing is that a specific compound has a specific demonstrated effect at a specific dose, and that chocolate is one vehicle among several for delivering it.

Not a claim about cardiovascular disease, blood pressure treatment or mortality.
Not a claim about the food as a whole, only about one compound.
Other reported benefits sit outside this authorisation and need their own evidence.

Four questions that separate substance from marketing.

How to read a chocolate health claim

Ask which compound is being credited, at what dose, for what measured effect, and whether the product in your hand actually delivers it. A claim that survives all four is worth taking seriously. Most front-of-pack chocolate health language fails at the third or fourth question: it names a compound and a benefit without a dose, or states a dose the product cannot be shown to meet.

Treat antioxidant language with particular care. Antioxidant capacity measured in a test tube is not the same as an effect in a person, and regulators in several jurisdictions have restricted general antioxidant claims for exactly that reason. The authorised cocoa flavanol claim is notable precisely because it is specific enough to be checked.

Which compound, at what daily dose, for what measured endpoint?
Does this product declare a measured content, or only imply one?
General antioxidant claims are weaker than a specific authorised claim.

Keep beside the wrapper

The authorised claim, and the things it is routinely mistaken for

Statement	Status	Why
Cocoa flavanols help maintain the elasticity of blood vessels, contributing to normal blood flow	Authorised in the EU	Cause and effect established for endothelium-dependent vasodilation at 200 mg daily
Dark chocolate is good for your heart	Not this claim	Broader outcome, different evidence base, not what was authorised
Dark chocolate lowers blood pressure	Not this claim	A treatment claim; outside the scope of the authorisation
Dark chocolate is high in antioxidants	Weak as stated	Test-tube antioxidant capacity is not an established effect in a person
Any 70%+ bar delivers the benefit	Not supportable	The condition requires high-flavanol cocoa, which percentage does not indicate

Three durable ideas

Leave with a model,
not a slogan.

01One narrow claim is authorised, for one measured effect, at one specified dose.
02Cacao percentage is not a proxy for flavanol content, and processing reduces it.
03A claim about a compound is not a claim about the food it arrives in.

Evidence used · reviewed 25 July 2026

Cocoa flavanols and maintenance of normal endothelium-dependent vasodilationEuropean Food Safety Authority · Public institution↗FoodData CentralUnited States Department of Agriculture · Official data↗Directive 2000/36/EC relating to cocoa and chocolate productsEuropean Union · Standard↗Codex standards for cocoa products and chocolateFAO / WHO Codex Alimentarius · Standard↗

Asked and answered

The questions
people actually type.

Short answers to the follow-ups this page raises. Each stays inside what the evidence above supports.

01Is dark chocolate actually good for you?: One narrow effect is authorised in the EU: cocoa flavanols help maintain the elasticity of blood vessels, contributing to normal blood flow, at 200 mg of cocoa flavanols daily. That is a real, substantiated physiological effect. It is not the same as a finding that dark chocolate prevents disease, lowers blood pressure as a treatment or extends life, none of which this authorisation covers.
02How much dark chocolate would I need to eat?: The EFSA opinion described the 200 mg daily condition as obtainable from about 10 g of high-flavanol dark chocolate or about 2.5 g of high-flavanol cocoa powder. The catch is the words high-flavanol: unless a product declares a measured flavanol content, there is no way to know whether it reaches that level, and most bars do not declare one.
03Does a higher cacao percentage mean more flavanols?: Not reliably. Percentage describes the share of cacao-derived material, while flavanol content depends heavily on processing — fermentation, roasting and especially alkalization all reduce it. An 85 per cent bar made from heavily processed cocoa can carry fewer flavanols than a lower-percentage product made to retain them.
04Is Dutch-process cocoa lower in flavanols?: Alkalization is the processing step most associated with flavanol reduction, because the treatment that softens acidity and darkens colour also degrades these compounds. If flavanol content is the reason you are buying cocoa powder, natural rather than alkalized is the more consistent choice, though only a declared measured figure settles it.