When can we say a wine is sweet? What is sweetness in a wine?

Training on Detecting the Sweet Flavour at the Tasting Panel of Vila Viniteca

When working on the basic “sweet” flavour with our panellists, we searched for the role of sweetness in the wine. During this process, many doubts came up: When can we say a wine is sweet? Where is the boundary between a sweet and a dry wine? What kind of paper shows the acidity in the perception of sweetness in a wine? And do the bouquets intervene in its perception?

Our Tasting Panel evaluates wine sweetness according to its intensity by means of a linear scale. For training purposes, there were created reference patterns of every intensity zones on a scale from 0 (null perception) to 10 (clearly identified and intense perception). The challenge and the goal were to achieve patterns that were the most representative and most faithful to reality in this area. In order to develop such patterns, we were guided by different hypotheses based on our knowledge and experience with trial-and-error tests.

Where does the sweetness of a wine come from?

Sweetness is one of the essential tastes or flavours which first develop at birth and one of the most important in the Western culture. According to the Dictionary of the Spanish Language edited by the Royal Spanish Academy (RAE), sweetness “is something that causes a kind of a specific soft and pleasant sensation on the palate, similar to that produced by honey or sugar”. But, why wines can be sweet? In general, it can be for two reasons, the first one  is that wine is made from grapes which, as any other fruit, have high levels of sugar (glucose+fructose and sacarose). In some cases, after the fermentation (process in which sugars are transformed in alcohol thanks to yeast action) there can still remain sugars in the wine. They are known as depleted sugars or, more common, residual sugars. The second cause is due to the addition of sugar to the wine, a more common practice for the sparkling wines and some of sweet types. Both are responsible for the sweet sensation done by the wine. 

It is important to notice that our sweetness scale is focused on the dry wines, we mean, it doesn´t include fortified sweet wines or naturally sweet. In this manner, it allows us to obtain more faithful or adjusted data.

Regulatory a “dry” wine is that which has equal or lees than 4g/l or residual sugar (for more information check 3.1.6. of International Norm for Wine Labelling from IOV. This rule is not according to our way of understanding the concept of dry wines, due to the fact that other sensorial factors may intervene in the general perception of a wine sweetness.

In the same way, we think that white and red wines cannot be evaluated with a same intensity scale, its composition makes them very different and as a result, it is difficult to evaluate them according to the same intensity parameter. Once established these premises, we start to elaborate the reference patterns.

We explain how we develop the reference patterns for each one of the sweetness intensity scale, starting from our hypothesis and from the obtained conclusions after each test.


White wines

1st Hypothesis: Real wines with a known sugar content (g/l)
We create a sweetness scale according to known sugar concentration with real wines and available for sale.

First used patterns


The results were quite satisfactory. We confirm that wine acidity plays a really important role in the sweetness perception, higher acidity less final sweetness perception.

The main challenge was to train the panellists to differentiate well the texture and flavour sensations, regarding sweet perception, because it was creating confusion.

These patterns didn’t have reproducibility, when the vintage was changing, stopped having the same characteristics. For this reason we kept searching for patterns which were satisfying our expectations and that mainly were easy to reproduce and stable.

2nd Hypotheses: Modification of wine sugar concentration (g/l)

From a white base wine –Xarel·lo 2017 (Catalunya)–,  with 0,5 g/l of residual sugar we add different concentrations of RCM (Rectified Concentrated Must) to obtain the patterns for each scale zone that we wanted to train:


From the first sessions, the panellists gave quite unified results and we didn´t need too many sessions to train them, fact that proved the quality of used reference patterns.

Anyway, we decided that, in order to complement the training, we would use in particular moments some of the real wines which we used in the first hypotheses.

Red wines

1st Hypotheses: Real wines with different sugar concentrations (g/l)

Initially we keep the same first hypotheses for red wines as with white wines and we use real wines. We choose 5 different references to cover different zones from the chosen intensity scale according to the valuation of the technical team regarding the sweetness sensorial perception of the wines.

First patterns used:


At our internal checking wine tasting, previous to the Panel, we verify that wines have an ascendant sweetness perception, even if we note very few differences between some samples.

The Panel couldn’t identify the 3 average values from the scale. After different training sessions with reiterated difficulties, we discarded the patterns.

2nd Hypotheses: Sugar concentration modification (g/l) of the wine (I)

We use a red base wine – Garnacha, Samsó and Ull de llebre 2016 (Catalunya) with 0.61 g/l of residual sugar – and we modify it to obtain the following concentrations:

  • 2 g/l
  • 3 g/l
  • 4 g/l
  • 5 g/l
  • 10 g/l
  • 15 g/l
  • 25 g/l
  • 30 g/l


The first checking wine tastings done by the technical team showed that, with the used wine base, from concentration 5 g/l we had a perception similar to the sweet wine (of raisin of fortified grape).

Following this hypotheses, we decided to present to the 15 panellists the concentrations comprised between (o-61 g/l and 5 g/l) by means of a comparison trial in pairs – which consists in comparing simultaneously 2 samples with different sugar concentrations. The purpose was to see if they were able to distinguish amongst different concentrations.

The results were the following:

Table of obtained results in training session (21/02/2018)

*It is the certainty level (or probability), mentioned in percentage, with the aim of carrying out an estimation of a parameter by means of sample statistics

We noted that the Panel wasn’t able to give homogenous answers, as they distinguished between small differences (between the concentrations 0.61 and 2g/l), but not big (about 2 and 3 g/l or 2 an 4 g/l). After different sessions with these patterns, we concluded that the obtained results weren’t satisfactory. The panellists had problems to analyse sweet sensation due to the tannins and acidity.

Because of the observed difficulties, we decided to enlarge the intensity scale, which at the same time allowed us including wines that, even if in some cases overpassed the border which labelling legislation indicates by the mention “dry” (a maximum of 4g/l), we consider them included in dry wines group according to their sensorial perception. Our high part of the scale would be formed by wines which would be considered as sweet, and which for us would represent the border with the sweet wine scale. They would be wines with more or less 100 g/l of sugar.

3rd Hypotheses: Sugar concentration modification (g/l) of the wine (II)

Base wine: Monastrell 2013 (Yecla) which, compared to the first used base wine, had more body and alcohol. We modified it into the following concentrations:


After passing the samples to the Panel, the panellists didn’t see significant differences between low zone pattern and low-intermediate zone one from the scale, but they did with the pattern from the intermediate high zone. As a result, we decided to validate only the patterns with concentrations of 5 g/l and 10 g/l.

After several sessions, the panellists gave repetitive results and in accordance with the expected ones.

Due to the fact that the scale remained short with these two patterns, we decided to complement it with real wines which covered the low and high parts of the scale using the following references: for the low part a Nebbiolo of Langhe (Piemonte) from vintage 2014 and for the high part of the scale a Corvina, Rondinella and Oseleta 2012 of Valpolicella (Verona) from vintage 2012.


Definition of sweet concept in the wine world: at regulation level we don’t find the border between a sweet wine and a dry wine in organoleptic sense, that means that, starting from how many grams each litre of sugar (taking into account its acidity level, alcohol, etc.), we can say that a wine is sweet.

Alteration of the real perception of sweetness: when analyzing the sweetness of a wine, we realized that there are components that can significantly alter the sensorial perception of this characteristic. These components are:

  • Alcohol: in “high” graduations between 14% and 15% and when the alcohol is well integrated with the wine
  • The flavour: we verify that some wines flavours can influence the panellist, these are toasted caramel, fresh fruit or lollipop. An association error occurs; the person associates a specific aroma with a recurring taste, in this case sweet.
  • The greasy texture of wines that may have been aged on lees or with lower acidities increase the sensation of sweetness.

Red wines have more components than whites (tannins, etc.), making it difficult to make patterns from the modification of a base wine, since the integration of an external component is more complex, as well as its evaluation.


In order to avoid distraction to our senses when analyzing the sweetness of a wine, we proposed to our panelists to use a synchronized swimming clamp to cancel the sense of smell. We also created a tasting protocol to better analyze the sweet (see Annex 1), since we verified that, in most cases, the human tongue perceives the sweetness on the tip of the tongue and the first seconds in which the wine enters the palate are crucial in determining intensity.

We found that to develop patterns with red wines it is better to use a base wine that is rich in structure, moderately alcoholic and has low acidity. The MCR is better integrated and the perception is much more real.

The final reference patterns used to train the Tasting Panel were as follows:

Take into account that the linear intensity scale of sweetness for wines that our Panel has is its own. We have prepared it under our criteria and taking into account more parameters than the concentration of reducing sugars in the wine, such as acidity or texture. It has also been designed according to our criteria and perceptions within a group of people belonging to the same society, culture and customs.

Annex 1

Tasting protocol to evaluate samples of the SWEET characteristic

1. Bring the wine to your mouth with a generous gulp, analyzing if you perceive any flavour on the tip of the tongue.
2. In a second phase, distribute the wine over the entire palate, analyzing its texture, especially if it is unctuous or sticky. Attention, discern if it is caused by alcohol or a consequence of aging.
3. Before spitting the wine, we leave it immobile on the tongue so that it can cover all the taste buds and can gain temperature.
4. Then we will spit it out observing how the liquid leaves our palate paying attention to its viscosity when it is spit out.
5. When we no longer have wine in the mouth cavity, we will analyze if we have a sweet sensation on the side of the tongue and on the tip, and if this is accompanied by a sensation of satiety, with thicker saliva and a sticky mouth.
6. Then we will pass the tongue over the gums, teeth and the walls of the palate to check if a lubricating layer has been created that disappears with the action of our tongue, since in some cases the sweetness is responsible.

Hits: 1270

Leave a Comment

Your email address will not be published. Required fields are marked *