AMA: Ask Me Anything (About Cocoa or Chocolate) #9: Roasting Cocoa

AMA: Ask Me Anything (About Cocoa or Chocolate) #9: Roasting Cocoa

The Question

Ethan Swift of Island Sharks Chocolate, Hawai’i, asked, “Roasting criollo requires lower temperatures (than other cacaos) and same time right?”


Like many things in cocoa and chocolate, I wish it were that simple. 

In this case I think this is not actually the right question to be asking as it’s based on a flawed assumption: There is such a thing as an “optimum” roast profile for a cocoa bean. 

In my opinion there is no such thing as an optimum roast profile. Any single roast profile is a compromise among possible flavor outcomes influenced by roasting.

That said, there is a kernel of truth in the idea that Criollos benefit from rather than require a lighter hand in roasting. This gentler roast can be a combination of lower roasting temperature and/or shorter roasting time. All depends on the objectives of the chocolate maker. 

A good analogy here is roasting coffee. I can take the same green coffee bean and give it a breakfast (light) roast or an espresso/Italian/French (very dark) roast – with a wide range of possibilities in-between. The same beans with different roasts yields coffee that tastes different. In general, the darker the roast the more burnt and bitter the flavors of the coffee. Lighter roasts reveal other flavors that may be in the beans. Some varieties of coffee beans, robustas for example, have less genetic potential for delicate non-coffee (e.g., fruity and floral) flavors while other varieties, arabicas, have greater genetic potential – and there is a lot of variety for potential within each broad category.

Cocoa beans with white or predominantly white cotyledons are easier to over-roast, cooking out the interesting flavor precursor chemicals developed during fermentation and drying. In general, white beans (not all beans with predominantly white cotyledons are genetic Criollos – there are such thing as “albino Forasteros”) “should” be roasted less heavily. However – there may be a lack of complete information about the genetics. Some white bean cocoas are mixes of Criollo and Forastero genetics but be fermented as 100% Criollos, which means they are being fermented incorrectly. Carmelo from Tabasco is one example.) This would suggest a different approach to roasting is in order compared with a white bean cocoa that is properly fermented.

With cocoa, different combinations of time and temperature yield different flavor possibilities but the same general rules are in play here:

  • Some varieties of beans have more (although different might be a better way to characterize this) genetic potential for flavor than other varieties of beans, and this potential is first modified by harvest and post-harvest processing practices.
  • Lighter roasts reveal the flavor potential of beans certain ways.
  • Darker roasts reveal the flavor potential of beans in other ways.
  • Choosing any one combination of time and temperature is an aesthetic choice on the part of the chocolate maker – a decision about what they like and what they think is a good expression of the flavor potential of the beans. 
  • DIfferent chocolate makers will choose different combinations of time and temperature for the same beans, resulting in very different-tasting chocolate. Any one combination is not in intrinsically better than any other (although some combinations can be objectively ruinous), we’re talking highly subjective creative processes here.

Most chocolate makers I have spoken with have approaches to roasting that are based on their experience working with the same or similar beans, or they may have a general philosophy that informs their approach to roasting.

For example, some makers work very hard to make sure they never come close to over-roasting their beans. They deliberately choose lower temperatures and shorter times to avoid burning or roasting out flavors they find interesting. Often these decisions are made experientially: the beans are roasted until they “smell done.” There are basic temperature and time parameters but these can be over-ridden a la minute. This is a part of the art/craft of small batch specialty chocolate making. 

Very often, many of these makers will not even experiment with heavier roasts to they are not even aware of the full flavor potential of their beans. In my experience, flavor development during roasting is not a linear process. Different combinations of time and temperature deliver different flavors and flavors may disappear at one time/temperature combination and be present in another.

Very often the decisions are made when the senses are at their least acute. If the maker has been smelling the roasting beans for 20 minutes that extended contact with the smell of roasting will change their perception of the smell. If the beans are tasted at different temperatures the perception of their taste will change as well.

A Different Approach to Roasting

My approach to roasting is based on my firm commitment to blending. If there is no one optimum roast profile (my contention) you can use roast blending to create chocolates that combine the desired aspects of different roasts. A light roast might have a more delicate floral/fruity quality and a heavier roast might have some very nice nut notes. In an intermediate roast these flavors might be out of balance, unpleasant, or not present. By blending two or more roasts in different ratios it’s possible to make more interesting-tasting chocolate while at the same time still making a single-origin chocolate if that’s what’s desired. It is also easier to hit the same flavor profile more consistently as that can be accomplished in the blending step, not the roasting step.

What I advocate is first roasting the beans and making liquor and chocolate according to an accepted standard. Cocoa of Excellence protocols come to mind. This way you have a reference against which to evaluate future tests.

I then advocate for making a grid with times on one axis and temperatures on the other. Beans are roasted and sampled at each combination of time and temperature combination. The beans are allowed to cool to the same temperature for evaluation. The evaluation will yield time and temperature combinations that give tasty results where the senses are not influenced by environmental factors. These time and temperature combinations are then used to roast larger batches which are made into liquor and chocolate and evaluated against the reference batch. 

From there, informal blending can happen, tasting different combinations to see how mixing them influences the flavor of the combination. Round of this are used to hone in on specific ratios.

For the real geeks out there (and if you got this far you are probably a real geek), it’s also fun to note that grinding and refining nibs at different roast levels will yield different results than blending ready-made chocolates in a tempering machine. DIfferent results – not necessarily better or worse. Different.

This same phenomenon holds true with different roasting methods. The same beans roasted in a 35kg drum roaster will yield different results at the same time/temperature combination than when roasted in a convection oven. Again, different results, not necessarily better or worse. It could be that blending roasts using different roasting methods can be interesting. And, in fact, this is something that Heinde & Verre do. I touched on their creative approaches to blending in this article


If you are interested in this topic at all I recommend you read the article and the comment below (which I copied over from LinkedIn). I also recommend you buy some H&V chocolate (all of them!) and taste the results for yourself. 


What are your thoughts about roasting and blending? Please share with the community in the comments below.


Media Inquires

If you are a member of the media and want to engage with me on any or all of these topics or need a source for any reporting on cocoa and/or chocolate, please send an email to

Archived Comments

Roasting is a matter of temperature over time. There is nothing to say that temperatures need to be constant over time and I agree that starting at a lower temperature and then increasing to a higher temperature or vice versa – important aspects of a more sophisticated roast profile – will change flavors in different directions.

But still, within the idea of varying temperature over time there is a useful distinction between lighter and heavier roasts. By restating the question I wanted to move away from the very common conception (to the point of mythology) that there is an “ideal” time and temperature for a particular “genetic variety” of cocoa and the goal is to find that ideal roast profile.

My position is that roasting is fundamentally a creative process that is amenable to different approaches and too many makers reject out of hand acknowledging or exploring these different approaches.

Over on LinkedIn, Heinde & Verre co-founder Ewald Rietberg commented:

It is not just the “amount” of roasting (heavy or low), but particularly the pattern of the roast profile that matters: eg starting low and ending high, constant at one temperature, or diminishing temperatures, a parabolic curve, two stage patterns, etc. At Heinde & Verre ( Instagram Heindeverre) we tend to distinguish patterns that reveal fruity aromas, patterns that reveal herbal and flowery aromas and patterns that reveal chocolaty, nutty and coffee aromas. The parameters (temperatures and time) may vary per type of cacao.

You've successfully subscribed to The Chocolate Life
Great! Next, complete checkout for full access to The Chocolate Life
Welcome back! You've successfully signed in.
Success! Your account is fully activated, you now have access to all content.
Success! Your billing info is updated.
Billing info update failed.