Roast in oolong tea is a heat treatment, separate from oxidation, which concerns changes managed while the leaves are being processed; Taiwan’s official processing account treats fixation, drying and roasting as separate operations. A label such as “medium roast” is the start of a description, not its answer: ask medium against which comparison, applied to what tea and when. Read a label under three headings — cultivar or material, oxidation and roast — and leave a blank where the seller says nothing.
When a packet says “medium roast,” ask what completes that description: medium against which comparison, applied to what tea and when? The label is the start of the explanation, not its answer.
Keep roast as one part of the tea’s making and one part of your preference. You might enjoy a pronounced toasted note in one sample and a quieter finish in another. Recording that difference is more useful than choosing one roast level for everything.
Traditional charcoal-roasting baskets and pits at Wang Tea in Taipei.
有記茶行/傳統炭焙屋/李宜蘭 — Leeilan54. CC BY-SA 4.0. Proportional resizing and WebP conversion; no crop or retouching. The original licence also applies to these derivatives.
Keep three pieces of information separate
Start a note with three headings: cultivar or material, oxidation, and roast. Copy what the seller actually states under each. Leave a blank where the information is missing.
A cultivar name identifies the cultivated plant material. Oxidation concerns changes managed while the leaves are being processed. Roast concerns a heat treatment, whose place in the production history matters. Taiwan’s official processing account distinguishes fixation, drying and roasting as separate operations with different purposes.[1]
The distinction is visible in experimental work. In a Tieguanyin study, tea made from the same fresh material received the same earlier processing, while the final roasting treatment varied.[2] The roast was a variable that could be changed without starting with a different named tea.
When a listing says only “light”, ask which heading it belongs under. A lightly roasted tea, a lightly oxidised tea and a cup you find light in flavour are three different descriptions.
Aroma does not move along a single scale
In the Tieguanyin experiment, longer roasting reduced the panel’s floral-aroma scores and increased roasted-aroma scores. A higher-temperature finishing treatment intensified the changes.[2]
A separate study of Da Hong Pao found a different pattern within its tested charcoal-baking conditions: grassy character diminished and floral-fruity character became more prominent as baking continued.[3]
Together, these findings are a reason to ask about the particular tea. They do not support a rule that roasting always removes flowers, always creates sweetness or always improves quality. The starting material and the treatment belong in the explanation.
Write the aromas you notice before deciding whether the roast is good or bad. Toast, nuts, fruit, flowers or wood can be prompts, but use them only when they fit the cup. “I cannot separate these aromas yet” is an honest note too.
Ask what the roast label means for this batch
You do not need every minute of a producer’s heat programme. A few details make a comparison much easier:
Is the roast description the producer’s wording or the retailer’s interpretation?
Is this the initial finished tea, or has it received a later roast?
When was the most recent roast?
Is another roast level available from the same starting batch?
What difference does the maker intend you to notice?
Treat “charcoal roasted” as a description to investigate rather than an automatic ranking. Ask about the maker and the batch. Likewise, a smoky impression in the cup is something to describe; it is not, by itself, a complete record of how the tea was heated.
If the comparison uses different teas, list the other differences alongside roast: cultivar, harvest, origin and storage. You can enjoy the tasting while leaving the cause of a difference unresolved.
Make a roast comparison with a narrow question
The cleanest practical comparison is two roast treatments of the same starting tea, if a producer can supply them. Otherwise, choose two clearly identified samples and be honest about the differences. This is an editorial tasting exercise, not a laboratory test or a personally tested recipe.
Use a brewing method you already handle comfortably. For the first comparison, keep the water, vessels, leaf weight and timing the same and write them down. That gives you a starting comparison; it does not guarantee that both teas are at their best.
Before looking at the supplier’s tasting notes, write three short lines for each cup:
First aroma: what is easiest to notice?
Balance: does that impression leave room for other aromas and tastes?
Finish: what remains, and do you want another sip?
Let both cups cool to a comfortable drinking temperature before tasting. Revisit them as they cool further. If you use repeated infusions, compare corresponding infusions rather than one tea’s first cup with the other’s third.
In a later session, adjust each tea on its own. If one was too concentrated, try gentler extraction before blaming the roast. Keep what you observed separate from your explanation of why it happened.
Resting is a question for the particular tea
If a maker recommends waiting after roasting, ask for a batch-specific explanation and storage instructions. Neither of the experiments above establishes a compulsory resting period for every oolong.[2,3] A fixed number of weeks cannot account for every starting tea and finishing treatment.
You can make a modest observation of your own: record the opening date, keep the remaining tea under the recommended conditions, and taste it again later using the same method. Describe the change without assuming it proves a universal ageing rule. Do not leave the packet open in a humid room to hurry the process along.
Buy the description you can use again
Keep the batch, the latest roast date if known and one plain sentence about your preference. “I liked the toasted aroma but preferred a shorter first infusion” is more useful for the next order than “heavy roast is best”.
Keep the conclusion attached to the tea: which part of this roast did you enjoy, and under which brewing conditions? That record can help you reorder a favourite while leaving room for a different sample to change your preference.
Personal experience — Pavel Rachkouski, founder. One person’s notes, not advice.
When I sort teas for myself, GABA and the soft oolongs go next to green tea, as the lighter, gentler ones. On the other side are the pu-erhs, darker and heavier, which I half-jokingly call the strong, heavy teas.
Frequently asked questions
Is roast the same as oxidation in oolong?
No. Oxidation concerns changes managed while the leaves are being processed; roast is a heat treatment with its own place in the production history. A roasted tea, a lightly oxidised tea and a cup you find light in flavour are three different descriptions.
What does a roasted oolong taste like?
It depends on the tea and the roast. In a Tieguanyin experiment cited on this page, longer roasting reduced floral-aroma scores and increased roasted-aroma scores. Compare two samples and write down what you notice.
What does “medium roast” mean on a packet?
Treat it as the start of a description: ask medium against which comparison, applied to what tea and when.
Cao, Q.-Q., and colleagues. Sensory and chemical characteristics of Tieguanyin oolong tea after roasting. Food Chemistry: X, 2021, 12, 100178. Controlled roasting treatments and sensory results, especially sections 2.1 and 3.1. Access note: this account uses accessible indexed passages from the authors’ Materials, Results and conclusion. The complete methods and conflict-of-interest declaration were not available for this editorial check; this is not a full-paper review.