How to Choose Good Chocolate

Most people choose chocolate using one or more of four signals: cocoa percentage, packaging (colours, words, images, fancy claims), a familiar brand, or certifications. I have judged at the International Chocolate Awards and tasted thousands of bars, and those four signals miss most of what matters.

I get this question all the time, in workshops, in shops, from friends: how do I choose better chocolate? The shelf is loud, the claims overlap, and the wrapper of a thoughtful craft bar can look nearly identical to the wrapper of an industrial confectionery. This is the answer I rarely have time to give in person.

How to tell if chocolate is high quality (and how to spot a real craft bar)

Short version: turn the bar over. High-quality chocolate tells you where the cacao grew, lists cocoa butter as the only added fat, keeps the ingredient list short, and skips vanillin, PGPR and alkalised cocoa. A real craft bar names a region, farm or harvest; an industrial imitation hides behind a country-of-manufacture claim like “Belgian chocolate”. These checks pick the bars worth tasting; tasting them is how you find out which are good. Everything below is the long version.

How to read a chocolate label: start with the ingredient order

Ingredients sit on the back of the wrapper in descending order of weight. That gives you the skeleton of any bar in a single glance.

For a plain dark bar, look for cacao first, then sugar, sometimes cocoa butter. Two and three ingredient bars are common, four is still fine, and past five or six in a plain bar you have drifted into confectionery, where the recipe does the talking instead of the cacao.

Milk and white chocolate work differently. Milk powder, sugar, and cocoa butter are all doing real work, so five ingredients is normal. Sugar first in a milk bar usually just tells you the product is more sweet than chocolatey.

What cocoa percentage really means (and why it is not quality)

More cacao means more intense, less sweet, and often more revealing, which is a different thing from better.

Cocoa percentage tells you how much of the bar is cacao by weight, mass and butter combined. The rest is mostly sugar (and in milk bars, milk solids). A 70% dark bar is roughly 30% non-cacao. In a two-ingredient bar, that means 30% sugar. A 70% with added cocoa butter can taste smoother than another 70% without it, because the butter dilutes the bitterness.

I would rather have Heinde & Verre’s 71% Pristine Nativo than most 85% bars I have tasted. The 71% knows what it is. The 85% often does not. The percentage predicts sweetness and intensity; quality lives elsewhere on the label.

If you want to see how this plays out across the catalog, the cocoa percentage chart lays out every band side by side, from sweet milk chocolate up to 100% chocolate. Each band page explains what to expect and links to real bars at that strength.

Fat: cocoa butter and its substitutes

Cocoa butter is the reason chocolate is not just “cocoa plus sugar”: the fat phase that gives a bar its snap, gloss, and the way it melts on your tongue and releases aroma. Replace it with cheaper fats and you have changed the category, even if the wrapper still says chocolate.

EU labelling rules allow up to 5% of certain vegetable fats in chocolate, and a bar that uses them must say “contains vegetable fats in addition to cocoa butter” on the label. For a quality-focused choice, the rule is simple: in a plain bar, any added fat besides cocoa butter counts against it.

Emulsifiers: not all the same

Emulsifiers help fat and cocoa solids stay dispersed, and they also help a bar flow through industrial equipment. Which one a maker reaches for, and whether there is one at all, fall on a hierarchy.

Plenty of decent makers use a small amount of lecithin for flow, so lecithin alone is a weak signal. The combination to act on is PGPR plus a long ingredient list plus a vague origin, which is industrial confectionery telling you what it is.

Flavouring: cacao should not need perfume

Cacao carries hundreds of aroma compounds, and a careful single-origin bar tastes like itself without help, so the best ingredient list for a plain dark bar names no flavouring at all. Real vanilla, as a pod, a vanilla extract or “natural vanilla”, is acceptable rather than a warning sign, and you will meet it a little more often in milk bars and inclusion bars.

Vanillin (synthetic), “artificial flavour” and vague “flavouring” are a different matter, because in a plain dark bar their job is masking rather than seasoning.

Single origin chocolate: country is good, specificity is better

Origin language is where wrapper marketing and wrapper quality diverge. “Belgian chocolate” or “Swiss chocolate” names a manufacturing style in a country that mostly does not grow cacao, so it leaves the only question that matters here, where the bean grew, completely unanswered.

The better hierarchy looks like this:

On Chof you can browse bars by named origin or cacao variety to see how this plays out, from the bright red fruit of Madagascar to the floral Nacional of Ecuador. Specificity makes a good bar more likely without guaranteeing one: makers confident in their cacao are more willing to name where it comes from.

Bean-to-bar chocolate (and tree-to-bar): what the label really tells you

Bean-to-bar means the maker controls the transformation from raw bean to finished bar in their own workshop. Tree-to-bar adds the farm side, with a maker who grows the cacao or is tied to a single plantation.

Bean-to-bar is a statement about control: it tells you the maker owns the result, and some of those results are forgettable. What even the forgettable ones get right is disclosure, which still puts them ahead of most industrial chocolate on that count.

Alkalisation and Dutch processed cocoa: a quiet warning sign

Alkalisation (also called Dutching) treats cocoa with an alkaline solution. It reduces acidity and bitterness, darkens the colour, and makes the taste more uniform.

That is useful in cocoa powder for baking, where predictable colour and a softer flavour are exactly the point, and unwelcome in a plain chocolate bar, where it flattens the origin character you are paying for and suggests a recipe tuned for shelf appeal.

Less than 1% of the bars on Chof are marked alkalised. That count comes from label text, so it catches American labels, which spell the treatment out, far better than European ones. The wording to look for:

• “cocoa processed with alkali”
• “alkalised cocoa” or “alkalized cocoa”, and on Dutch and German labels “gealkaliseerd” or “alkalisiert”
• “Dutch processed” or “Dutched”
• “acidity regulator: potassium carbonate”, or sodium carbonate, E500 or E501

European labels vary more: some print the word, some name the salt, and some print neither, because EU labelling rules (Article 20) let an additive used on one ingredient, such as the alkali on a cocoa mass the maker bought in, go unlisted when it does no job in the finished bar. Read a European list for the salts as well as the words, and take a list without them as the absence of a warning rather than proof of none.

On Chof, alkalisation is one of the negative signals Chof Score picks up under Process.

What Chof Score actually looks at

Chof Score is an evidence-based quality signal that orders bars on Chof — a structural reading of how serious a bar is about the signals that usually predict good chocolate, rather than a verdict on taste. Four pillars do most of the work.

Two safety nets sit on top. Bars with substitute fats or identity failures (a “white chocolate” without cocoa butter, for example) get capped. So do bars with credible safety failures, like heavy metal limits exceeded.

How much evidence we have on a bar, and how consistent that evidence is, also shapes the final number. A bar with strong origin disclosure but no information on how it was made will not be scored as if everything were known.

Chof Score in action: Top 20 Chocolate Bars, ranked by what their wrappers and disclosures actually say rather than by popularity or reviews.

Same percentage, different bar

Both of these are dark bars of about 70% made with cocoa, sugar, and cocoa butter, with no PGPR, no vegetable fat, no vanillin and no alkalisation, and on the front they look almost identical. Chof Score reads what each maker discloses about ingredients, origin and process, which is a separate question from which bar tastes better, and on that reading it puts Heinde & Verre higher.

Both bars are good, and both pass every check on this page. The difference is what Heinde & Verre tells you that Vivani does not: the region and the named growers behind the cacao. Disclosure is what Chof Score reads, and flavour is a separate matter, since the same beans in four makers’ hands give four different bars.

What the wrapper never tells you

A bar can pass every check on this page and still taste flat, because the wrapper only records what the maker was willing to disclose, while tasting reports what actually survived the maker’s roast.

In International Chocolate Awards flights the samples are anonymised, and the same faults come up again and again: fermentation character (slaty, vinegary, smoky), texture defects (sandy, gritty), a roast that flattens the bean, and a finish that disappears too fast. None of them shows up on a label.

The wrapper checks above help you avoid the worst. Your mouth tells you the rest.

Real bars that show this in practice

Real bars from the Chof app, picked because they illustrate the principles above: high Chof Scores combined with diversity of category, country, and accessibility, so that you can actually find some of them. The number on each card is the bar’s current Chof Score.

Good supermarket-accessible signals

These are easier to find in larger grocers, organic shops, and specialty supermarkets in the Netherlands and the wider EU. They are proof that ingredient hygiene and origin specificity are possible at this level, even if the rarest and highest-scoring bars live elsewhere.

Specialty bars worth tasting

The kind of bars Chof was built around: short ingredient lists, named regions or farms, real cocoa butter, no alkalisation, and detailed process disclosure. Taste one of these next to a supermarket dark chocolate bar at the same percentage: the fruit, the acidity and the length of the finish are what you are paying the difference for.

Dutch makers worth knowing

If you are reading this from the Netherlands, three makers based here are consistently strong on every criterion above. I am in the Netherlands too, so worth a little extra attention. :)

Wrapper red flags

You will see this label, or something close, on a lot of bars in the confectionery aisle:

Translated: a sweet confectionery base with chocolate character, built on sugar first, substitute fats, PGPR, vague flavouring and no stated origin. There is nothing wrong with eating it. It is simply a different product from a bar that is trying to show you a place.

Frequently asked questions

A note on Chof Score methodology

The example bars come from the Chof app, selected by combining Chof Score with category diversity, supermarket accessibility, and country balance. Scores on this page are read live from the catalogue, so they move when a bar’s evidence changes. Chof is independent of any maker.

Chof is built and maintained by Felipe van Beetz, who holds the IICCT Level 2 Certificate in Chocolate Tasting and has judged at the International Chocolate Awards.

Drafting and research supported by AI tools; sources and citations checked by Felipe. How Chof writes its guides.

Further reading on Chof