Cider science

Perry chemistry, and what is actually different

How is perry chemically different from cider?

In short

In four ways that hold up. Pear juice carries substantial sorbitol, which yeast does not ferment, so a perry taken to dryness still tastes sweet. It carries citric acid alongside malic, which apple juice essentially does not. It is much poorer in the nitrogen a yeast can use, and much richer in proline, which a yeast cannot. And its tannin tends to be perceived as a drying grip rather than as bitterness.

The nitrogen difference is the one with practical consequences a maker meets on the first ferment, and it is now measured rather than asserted: a survey of thirty-eight Three Counties varieties found free amino nitrogen from 70 milligrams per litre down to 1.

Two other claims circulate widely and are not supported at the resolution they are usually stated. That perry is inherently lower in alcohol than cider is a statement about how perry is made rather than about pear juice, and that perry is unstable because of its sugars confuses two separate mechanisms.

Eight perry pear varieties measured in the Three Counties. The solid bars are assimilable nitrogen; the dashed bars are proline, which yeast cannot take up without oxygen. They do not move together.

Described in full

Shape: Eight named rows, each with two horizontal bars. The solid bar on the left is free amino nitrogen in milligrams of nitrogen per litre, on a scale to 80. The dashed bar on the right is proline in milligrams per litre, on a wider scale to 360. Each bar carries its figure as a number beside it.
The figures: Green Roller 70.6 and 133.4; Felix 69.7 and 164.3; Betty Prosser 49.2 and 269.9; Barland 24.8 and 145.6; Butt 13.3 and 17.9; Moorcroft 6.3 and 123.1; Red Pear 3.7 and 42.8; Tumper 1.0 and 337.2. All are single samples from a survey of Three Counties fruit taken between 2022 and 2024.
Why proline is separated: Proline is an amino acid, and any measurement of total amino acids counts it. A yeast fermenting anaerobically cannot assimilate it, because the enzyme that opens the proline ring requires molecular oxygen. So a juice can be rich in amino acids and starved of usable nitrogen at the same time.
Tumper: The extreme case, and the reason the figure exists. One milligram per litre of usable nitrogen beside 337 of proline — a ratio of more than three hundred to one. A total amino acid figure would describe that juice as well supplied. It is not, and a ferment on it alone has effectively nothing to eat.
The two do not track: Butt is low in both. Moorcroft is low in one and high in the other. Green Roller is high in both. There is no relationship visible here that would let a maker infer one figure from the other, which is why the survey reports them separately and why this figure does.
Scope: Eight varieties out of thirty-eight surveyed, chosen to span the range rather than sampled at random. Every figure is one juice from one collection in one autumn: the same survey found the same variety differing markedly between years and between tree types.

Sorbitol: the difference that is real and is often explained wrongly

Pears accumulate sorbitol, a sugar alcohol, in quantities apples do not. Saccharomyces cerevisiae does not ferment it under normal cider conditions, so it survives the ferment intact and is still there in the finished perry, where it tastes sweet. This is why a perry fermented to complete dryness — no residual fermentable sugar at all — can taste distinctly sweeter than a cider in the same state.

The mechanism is worth stating precisely, because the usual version gets it backwards. It is not that perry ferments incompletely. It is that part of what tastes sweet in pear juice was never fermentable, so the ferment finishing changes nothing about it. A hydrometer reading in a perry therefore overstates dryness relative to what a drinker will report, and the gap is a real one rather than a measurement error.

A modern survey observed the effect from the other direction. Four perry pear varieties were fermented singly under standardised conditions; three finished near 1.005 and Blakeney Red finished at 1.00898, much higher. The surveyor attributes the gap to unfermentable sorbitol and notes carefully that sorbitol was not directly measured in those juices. CiderHQ reports the attribution and the reservation together, because a hypothesis with its own author’s caveat attached is more useful than either half.

What is not knownCiderHQ holds no per-variety sorbitol analysis for any perry pear. The figures it carries are pooled pear-juice ranges from the general literature, marked as such on every record that shows one. The gap is large: sorbitol is the single most cited fact about perry and nobody CiderHQ can reach has published it variety by variety.

Nitrogen, and the proline trap

Yeast needs nitrogen in a form it can assimilate. In juice that means chiefly ammonium and the free amino acids, and the working measure is yeast assimilable nitrogen — or, where the ammonium fraction is not separately determined, free amino nitrogen. Cider apple juice is often short of it. Perry pear juice is very much shorter.

The trap is proline. Proline is an amino acid, and it counts towards any measurement of total amino acids; but the enzyme that opens its ring requires molecular oxygen, so a yeast fermenting anaerobically cannot use it at all. Pear juice routinely carries more proline than everything else put together. A juice reported as rich in amino acids can therefore be a juice with nothing usable in it.

The Three Counties survey makes this concrete. Tumper measured 7.8 milligrams per litre of total assimilable amino acids — one milligram per litre expressed as nitrogen — beside 337 milligrams per litre of proline. Red Pear measured 3.7 milligrams per litre of nitrogen. At the other end, Green Roller reached 70.6 and Felix 69.7. The spread between varieties is more than seventyfold, and nothing about the fruit tells a maker which end a variety sits at.

The consequence is that a perry ferment is much likelier than a cider ferment to slow, to finish late, or to make hydrogen sulphide, and that none of those outcomes reflects on the maker’s hygiene. It also means that blending perry is a fermentation decision as much as a sensory one: a blend can be assembled to taste right and still be a blend the yeast cannot finish.

Citric acid, and why it matters more than its quantity suggests

Pear juice carries citric acid alongside malic; apple juice carries essentially none. In quantity terms it is a minor component, and a total-acid figure reported as malic acid hides it entirely.

What makes it disproportionately important is that lactic acid bacteria metabolise citric acid, and among the products are diacetyl and acetic acid. So a perry that undergoes an unmanaged bacterial phase can develop both a buttery note and volatile acidity from a component the analysis did not report. This is one of the ordinary routes by which a perry becomes unstable in cask or bottle, and it has no equivalent in cider.

It also means the acid-and-tannin classification predicts the palate of a perry less reliably than it predicts the palate of a cider. The scheme was built for apples and it counts total acid; two pears with the same total can differ in what that acid is.

Tannin that reads as grip rather than as bitterness

A perry pear called bittersweet is often not bitter in the way that word suggests. Perry pear tannin tends to be perceived as a drying astringency across the whole mouth rather than as a taste at the back of the tongue, and the difference is one of chain length rather than quantity: bitterness is a receptor event that shorter procyanidins trigger, while astringency is a physical loss of lubrication that longer polymers cause by cross-linking salivary proteins.

The modern survey adds resolution here that no historic table could. It reports the polymerised “true tannin” fraction separately from the phenolic acids, and the two do not track. Butt measured 3 000 milligrams per litre of true tannin; Wasserbine measured 260 — more than tenfold apart in what the mouth will actually feel. Meanwhile Lullam carried nearly as much phenolic acid as tannin, an unusual split that means a juice which browns readily and tastes less astringent than its phenolic total suggests.

A historic perry table reports one “tannin” figure produced by permanganate titration, which responds to everything phenolic in the juice and merges these fractions. Comparing such a figure with a modern fractionated one is comparing two different measurements.

Two claims that do not survive the evidence

That perry is inherently weaker than cider. Perry pear juice gravities in the Long Ashton table run from 1.045 to 1.069, which spans and exceeds the ordinary cider apple range; Brandy at 1.069 and Tumper at 1.069 would ferment to a strength no cider apple beats. Where perry is weak it is usually because the ferment was deliberately arrested with sugar remaining, as the Domfront appellation requires, or because the fruit was picked early. That is a production fact, not a chemical one.
That perry is unstable because of its sugars. Two separate mechanisms get conflated here. Residual fermentable sugar in a sealed container is a refermentation risk and a genuine hazard; sorbitol is not fermentable by Saccharomyces and is not that risk. The instability perry is genuinely prone to comes from elsewhere — citric acid metabolism by lactic bacteria, and the low nitrogen that leaves ferments incomplete.

What would settle the open questions

Per-variety sorbitol analysis is the largest gap. Every general account of perry rests on sorbitol, and no source CiderHQ has reached publishes it variety by variety, so the site carries a pooled literature range and labels it as one.

Per-variety citric acid figures are the second. The acid analyses CiderHQ holds for perry pears — historic and modern alike — report total acidity as malic, which is the convention and which cannot distinguish the two acids.

And the nitrogen survey, valuable as it is, is one sampler collecting from one region across three autumns, with several varieties represented by a single juice. Where it sampled a variety twice the readings differ substantially. It is the best evidence available and it is not a settled body of numbers.

Also answered on this page

Questions this page covers, so you can tell at a glance whether it is the one you want.

How is perry different from cider chemically?
Why does dry perry taste sweet?
Why do perry ferments get stuck?
Does perry have less alcohol than cider?
What is proline and why does it matter in perry?

What people ask next

Questions readers ask about the things this page mentions. Each one goes to the section that answers it rather than to a page written to receive the question.

What is the difference between cider and perry — Cider is fermented apple juice and perry is fermented pear juice. Perry is lower in tannin, higher in unfermentable sorbitol so it tastes faintly sweet even when dry, and considerably harder to make.
How do i get rid of the sulphur smell in my cider — Racking with a little splashing usually blows off free hydrogen sulphide while it is still fresh. Once it has reacted into mercaptans the smell becomes rubbery and no longer responds to aeration.
What is yan and why does cider juice run short of it — Yeast assimilable nitrogen is the nitrogen yeast can actually use. Apple juice is usually short of it — in one Virginia survey of 108 samples, 94 per cent fell below the level wine practice treats as a minimum — which is why cider ferments stall more readily than wine ferments.
Why does cider need yeast nutrient
Why is perry not simply cider made from pears — Because the chemistry differs, not only the fruit. Pears carry sorbitol, which yeast cannot ferment, so a fully fermented perry can still taste sweet — and pear juice carries citric acid as well as malic, which a single total-acid figure does not distinguish.
Why is perry sweet even when it is fermented dry

Sources

What this page rests on. Where a source is marked as registered rather than read, CiderHQ is recording that the body is authoritative on the subject without claiming to have worked through the document itself. See our evidence policy for what each state means.

Perry Pears and Perry — Research and Survey Work
Cider and Perry Academy (Peter Mitchell) · specialist publication · passage verified 2026-08-25 · covers 2022–2024
Ongoing survey work by a cider and perry technologist who has taught production for decades, published openly and updated annually; retrieved and read in full on 2026-08-25. Thirty-eight single-variety perry pear juices analysed for gravity, titratable acidity, the tannin and phenolic acid fractions separately, and the individual amino acids by HPLC — the last of which nobody else has done at this scale for perry. Not peer-reviewed, and the author is explicit that the accompanying fermentation trials are initial surveys needing statistical treatment before differences are called significant; CiderHQ repeats that wherever the fermentation results are used.
Perry Pear Varieties from Long Ashton (NACM Short Report 8.2)
National Association of Cider Makers, Archive of Cider Pomology · research institute · passage verified 2026-08-25 · covers 1903–1963
A four-page compilation, dated March 2010, of the Long Ashton perry pear varieties: for each of twenty pears, its vigour, cropping habit, maturity window, specific gravity, acidity, tannin and an assessment of vintage quality. It rests on Luckwill and Pollard’s Perry Pears (1963), which CiderHQ has verified bibliographically and has not opened, so the figures are held at traditional-record tier rather than at research tier despite originating in institutional analysis. This is the report that ended a two-sprint gap: before it, no perry pear page on the site carried a cultivar-specific number.
The Science of Cidermaking and associated technical writing
Andrew Lea · reference work · passage verified 2026-08-24
Written by a food chemist who worked at Long Ashton on apple phenolics. Unusual among specialist cider writing in that it is primary-research-adjacent: the author is describing work he did, and cites the literature. This is why it is registered at tier 1 for chemistry while a general cider book is not.
Research on pear juice composition and sorbitol in perry
Various journals and institute reports · peer-reviewed literature · registered as competent for this subject
The evidence base for the single most important chemical difference between cider and perry: pears carry substantial sorbitol, which yeast does not ferment, so a fully fermented perry retains sweetness a fully fermented cider cannot.
National Perry Pear Centre
National Perry Pear Centre, Hartpury · specialist organisation · passage verified 2026-08-24
Read on 2026-08-24 — and the address CiderHQ had registered, perrypears.co.uk, is no longer the Centre at all. It fails the TLS handshake over https, and over plain http it resolves to a Sedo domain-parking page: an earlier pass in this sprint recorded it as retrieved on the strength of an HTTP 200, which was wrong, and the correction is the point. The Centre publishes at nationalperrypearcentre.org.uk. It is the closest thing British perry has to an identity authority: it holds at least two examples of most known varieties, states that the DNA of most of them has been determined and is used as an aid to identification, and publishes a variety register with synonym lists. Its 121-variety register is transcribed into the perry pear records, including the nineteen names it shows being claimed by more than one variety.