Do Multi-Volume Sets Sell for More?
By Antonio Lopes · · 12 min read
Everyone who inherits a matching set of books has the same instinct: this must be the valuable part. Twelve volumes, uniform bindings, gilt on the spines. It looks like money in a way a single paperback never does.
The received wisdom agrees. Sets are worth more complete than broken. Never split a set. A run of anything is worth more than its parts.
I run a database of 800,385 real auction and marketplace results, so I can check. The answer is not that sets are worthless — some are magnificent. It is that the premium is far smaller than anybody expects, and once you divide by the number of books, the picture completely inverts.
The headline: a set is worth about 7% more than a single book
Start with the crude version. Take every book lot whose description mentions volumes, delen, Bände, tomes or a complete set, and compare it with everything else.
Multi-volume lots type at €65. Everything else types at €61. That is a premium of 1.07× — seven per cent, on lots that contain several times as many physical books.
That is the whole received wisdom, gone in one line.
But a pooled comparison like that mixes different markets together, so I did it the careful way as well: same author, sets versus single volumes, only where there were enough of both. Across 87 authors who qualified, the typical set sells for 1.21× a single book by the same author — and for 28 of those 87 authors, sets sold for less.
So it is a real premium, and it is small. Roughly a fifth more money for several times as many books.
Now divide by the number of books
This is the part that changes how you should think about sets, and it is why the headline is misleading rather than merely disappointing.
I pulled the volume count out of the listing title wherever it was stated — "3 delen", "12 volumes", "5 Bände" — which gives 13,854 lots where I know both the price and how many books were in it. Then I looked at the total price and the price per volume side by side.
Read the blue bars and you would say sets do get more valuable as they grow: €60 for two volumes, up to €78 for twenty-one or more. A rise, but a gentle one — going from two books to twenty-one adds thirty per cent to the price.
Read the gold bars and the market looks completely different. Per volume, the price falls off a cliff: €30 a book in a two-volume set, €16 at four, €8.60 at seven to ten, €5 at eleven to twenty, and €2.19 once you pass twenty.
A twenty-six-volume encyclopaedia selling for €120 sounds like a substantial result. It is €4.60 a book. The typical single book lot in this database makes €61.
Within the same author, that comparison is even starker: sets sell for 1.21× a single volume in total, but 0.22× per volume. For 85 of the 87 authors tested, the per-volume price was lower than the single-book price. There is no meaningful exception.
What this actually means when you are deciding
Two practical consequences, and they point in opposite directions depending on which side of the transaction you are on.
If you are buying, big sets are the cheapest way to acquire books. Not the cheapest way to acquire value — the cheapest way to acquire physical volumes. At €2.19 a book, a twenty-volume run is priced as furniture. That is exactly why they are cheap: the buyer needs five feet of shelf and a way to move eighty kilos.
If you are selling, the set is a liability disguised as an asset. Every extra volume adds weight, packing cost and shipping risk, and adds almost nothing to the price. A seller with a twelve-volume set is doing several times the work of a seller with one book for the same money.
So should you break up a set? The arithmetic says yes far more often than the folklore admits — if the individual volumes have standalone appeal. Twelve separate lots at €25 each beats one lot at €75, and it is not close. The catch is that most volumes in a big uniform set do not have standalone appeal. Volume 7 of a general encyclopaedia is worth nothing on its own, which is precisely why the whole thing sells for €78.
The sets genuinely worth keeping intact are the ones where the value lives in the completeness itself: a plate book with the plates in all volumes, an author's collected works in a fine binding, a periodical run. And that turns out to be measurable.
"Complete" is worth saying, and worth checking
The word complete in a listing title is worth real money, and the effect grows with the size of the set.
| Volumes | Marked complete | Not marked | Difference |
|---|---|---|---|
| 2 | €60 (175 lots) | €60 (3,682 lots) | 1.00× |
| 3 | €70 (123 lots) | €61 (1,994 lots) | 1.15× |
| 4 | €75 (113 lots) | €65 (1,507 lots) | 1.15× |
| 5–6 | €95 (107 lots) | €70 (1,695 lots) | 1.36× |
| 7–10 | €95 (200 lots) | €70 (1,697 lots) | 1.35× |
| 11–20 | €94 (143 lots) | €71 (1,355 lots) | 1.32× |
| 21+ | €101 (79 lots) | €75 (980 lots) | 1.35× |
At two volumes the word does nothing — of course it doesn't, a buyer can see both books. From five volumes upward it is worth about a third more, consistently, across four separate size bands. Overall, sets described as complete type at €84 against €65 for those that are not (940 lots against 12,910).
Now the uncomfortable half of that finding. Out of 13,854 set lots, the number describing themselves as incomplete is four. Four.
Sellers do not volunteer that a set is broken. They leave the word out and let you assume. Which means "not marked complete" is not a neutral category — it is a mixture of complete sets whose sellers did not bother to say so and incomplete ones whose sellers preferred not to. If you are buying a large set and the listing does not say complete, count the volumes yourself before bidding. That habit is worth more than any figure in this article, and it is the same scepticism that reading an auction result properly calls for.
Old sets behave differently. Modern ones do not
The last thing worth testing is whether sets used to be a better proposition than they are now. They were.
For books printed before 1800, sets type at €110 against €78 for single volumes — a genuine 1.41× premium, from 1,794 set lots. In the 1800s it is €67 against €55, or 1.22× (2,613 lots). By 1900–1949 the premium is exactly nothing: €55 either way. From 1950 onward it is €63 against €60, which is noise.
Even here, though, the per-volume rule holds. A pre-1800 set makes €28.50 a volume, well under the €78 an individual pre-1800 book makes. Old sets are the best sets, and they are still worse per book than buying singly.
Why the era matters: before 1800, a multi-volume work was usually a single publishing project sold as one object — a natural history with plates distributed across the volumes, a folio atlas, a set of law reports. Losing a volume genuinely destroys it. After 1900, most sets are collected editions and encyclopaedias produced in large numbers, where the run is convenient rather than scarce. If you want the longer version of why the 1800s are the turning point for book values generally, that is does age make a book valuable.
The short version
- A multi-volume set typically sells for €65, against €61 for other book lots — a 7% premium. Within the same author it is 1.21×.
- The total price barely moves with size: €60 for two volumes, €78 for twenty-one or more.
- Price per volume collapses from €30 to €2.19. That is the number that decides whether a set is worth buying, shipping or breaking up.
- "Complete" in the title is worth about a third more on sets of five volumes or larger. Almost nobody writes "incomplete", so count the volumes yourself.
- Sets only outperform single books for material printed before 1900.
None of that makes sets bad. It makes them heavy. The value in a book is overwhelmingly determined by what the individual book is — its rarity, its condition, whose name is on it — and a set is just several books at once, which is why what a signature adds or a genuine early printing moves price far more than volume count ever will. The broader picture of what actually drives a book's value is in is my old book worth anything.
How this was worked out
Everything above comes from this site's database of 800,385 auction and marketplace results, reduced to the 562,778 lots that are genuinely books, maps or prints. The raw catalogue contains jewellery, model trains, records and cars, all of which have to be removed first — "lot with 6 albums" is not a book set, and leaving those in was quietly dragging the numbers down until I separated them.
That separation matters enough to spell out. Job lots — listings phrased as "lot with 8 books", a mixed pile sold together — are a different product from a matching multi-volume set, and they behave differently: 11,397 job lots type at €46, or about €10 an item. Merging them with real sets, as the obvious approach does, makes sets look worse than they are. Every set figure in this article excludes them.
Volume counts were read from the listing title, so this only covers sets whose seller said how many volumes there were. Prices are typical sales rather than averages — one €8,500 ten-volume topographical work would otherwise distort the whole 7–10 band. Sample sizes are given beside every figure so you can weigh them yourself.
To check what a specific set has actually sold for rather than relying on my averages, search the database — every result links back to the original listing, volume count and all.
