what does ‘extra-virgin’ really mean?
The phrase sounds like a promise of purity. It is in fact a technical specification with defined chemical limits and a compulsory taste panel — and the gap between the specification and what sits on a shelf is where the whole problem lives.
Extra virgin is not a marketing adjective, and it is not a claim about romance or hillsides. It is the top grade in a trade standard, and to carry the words a sample must clear a defined set of laboratory measurements and pass a formal sensory examination by a trained panel. That is a genuinely rigorous definition — considerably more rigorous than most food labelling.
The difficulty is not the definition. It is the distance between a sample tested at the mill and a bottle bought eighteen months later in a supermarket, and the fact that one of the required tests is a human judgement.
The tests, and what each one actually detects
Four families of measurement, plus a panel. Each catches something different, and none of them catches everything.
| Test | Limit for extra virgin | What it detects | What it misses |
|---|---|---|---|
| Free acidity | 0.8 g per 100 g, as oleic acid | Damaged, fermented or badly handled fruit; poor milling | Says nothing about taste or freshness after milling |
| Peroxide value | 20 milliequivalents of oxygen per kg | Early-stage oxidation — light, heat, air exposure | Can look fine in oil that is already old but past this stage |
| UV absorbance (K232, K270) | K232 up to 2.50; K270 up to 0.22 | Advanced oxidation, and the presence of refined oil | Modern low-temperature deodorising can slip under it |
| Ethyl esters | 35 mg per kg | Fermented fruit — olives stored in heaps before milling | Nothing about oxidation or blending |
| Sensory panel | Median defect zero, fruitiness above zero | Rancidity, fustiness, mustiness, winey and muddy defects | It is a human judgement, and panels can disagree |
The panel is the strong point and the weak point
The sensory examination is the only test that assesses whether the oil is actually pleasant. A trained panel of tasters scores defects and fruitiness under controlled conditions, from coloured glasses so that appearance cannot influence the verdict. For a bottle to be extra virgin, the median score for defects must be zero. Not low. Zero.
That is an admirable standard and it is where most disputed samples fail. It is also, inevitably, the least reproducible part of the system. Panels are made up of people, trained to a common method but not identical in sensitivity, and borderline samples can genuinely receive different verdicts from different accredited panels. Producers complain about this and they are not entirely wrong. But the alternative — grading purely by chemistry — would let through oil that is chemically clean and tastes of nothing at all, which is precisely the outcome the panel exists to prevent.
The fraud the tests were not built for
Crude adulteration — cutting olive oil with cheap seed oil — is comparatively easy to detect, because the fatty acid profile gives it away immediately. That is not where the sophisticated money is.
The serious problem is deodorised oil. Take lampante oil, the low grade that fails on taste, and treat it in a mild vacuum process at a temperature low enough that it does not leave the classic markers of full industrial refining. The defects come off. The oil becomes bland but not obviously faulty, and it can then be blended with a little genuinely good oil and presented as extra virgin. It may pass the standard chemistry, and a panel faced with a bland, defect-free blend has little to object to.
Tests exist that do catch this — markers such as pyropheophytins and the ratio of diacylglycerols, which track both thermal treatment and age. Some producing countries adopted them; the main international standard has not made them compulsory. That single regulatory gap explains a great deal about why the category remains vulnerable, and we have written more about the mechanics of it in how olive oil gets cut and the extra virgin that is not.
| What the grade means | Mechanically extracted, unrefined olive oil meeting defined chemical limits and passing a sensory panel with zero median defect |
|---|---|
| Free acidity | No more than 0.8 g per 100 g, expressed as oleic acid |
| Peroxide value | No more than 20 milliequivalents of active oxygen per kg |
| UV absorbance | K232 no more than 2.50; K270 no more than 0.22 |
| Ethyl esters | No more than 35 mg per kg in the EU |
| Sensory requirement | Median of defects zero; median of fruity attribute above zero |
| Not covered by the grade | Age at purchase, storage conditions after bottling, and low-temperature deodorising |
What a buyer can actually do
You cannot run a chromatograph in a supermarket aisle. But grading failure correlates strongly with age and careless handling, and those you can see.
- Buy by harvest date, not best-before. An oil more than eighteen months from harvest may be legally saleable and still taste of nothing.
- Buy dark glass or tin and refuse clear bottles that have been standing under shop lighting.
- Learn the defects. Rancid smells like old nuts or crayons; fusty is a sour, sweaty barnyard note from heaped fruit; muddy is a stale, dank sediment smell. Twenty minutes learning these is worth more than any label.
- Bitterness and pepper are not defects — they are the signature of fresh, phenol-rich oil. A cough at the back of the throat is a good sign.
- Be suspicious of a bland, faultless, cheap extra virgin. Blandness is what deodorising produces.
- Prefer a named origin and a named variety. Each one narrows the room for substitution.
Extra virgin: common questions
What does extra virgin legally mean?
That the oil was extracted mechanically without refining, meets defined chemical limits — free acidity no more than 0.8%, peroxide value no more than 20, and set UV absorbance and ethyl ester ceilings — and passed a trained sensory panel with a median defect score of zero.
Is extra virgin olive oil often mislabelled?
Category fraud is a documented and persistent problem, and shelf-aged oil frequently no longer meets the grade even when it did at bottling. The most difficult fraud to detect is mild deodorising of low-grade oil.
Why is the taste panel part of the legal test?
Because chemistry alone cannot tell whether an oil is pleasant. A chemically compliant oil can still taste rancid or fusty, so a trained panel scores defects and fruitiness under controlled conditions.
Does a low acidity number mean a better oil?
Not really. Free acidity is a threshold, not a quality scale. It tells you the fruit was sound and well handled at milling; it says nothing about how the oil tastes or how old it now is.
Can extra virgin olive oil stop being extra virgin?
Effectively yes. Grade is measured at a moment in time. Light, heat, air and age degrade oil, so a bottle that qualified at bottling may no longer meet the sensory or chemical requirements a year or two later.
Here is the uncomfortable thing about this grade: it describes the oil on the day it was tested, and almost nothing on the label tells you how far the bottle in your hand has travelled from that day. A great deal of oil sold entirely legally as extra virgin would not pass the panel if it were retested at the point of sale — not because anyone committed fraud, but because time, light and warmth do to olive oil exactly what they do to everything else. If the industry made one change, it should not be another chemical test. It should be a compulsory harvest date on the front label. Every serious producer would welcome it, which tells you precisely why it has not happened.
Chemical and sensory limits follow the International Olive Council trade standard and the corresponding EU regulation; detection-marker discussion reflects published analytical research.