In short: Almost every collection is worth what a few dozen cards are worth, and the rest is priced by weight. Sort the box into hits, playable rares, and bulk. Price a sample of the hits against sold listings, apply that average across the pile, and quote the other two piles per card. The 3,000-card box worked through below lands at about $430, and $372 of it sits in 40 cards.
Sort the box before you price anything
Tip the box out. Work in three passes and look up nothing yet, because pricing while sorting is how an afternoon turns into a fortnight. The first pass pulls anything shiny: a holo pattern across the artwork or the border, full-art cards, gold and textured edges. The second pulls anything old, which for Pokémon means a copyright line reading 1999 to 2003, a Wizards of the Coast credit, or a 1st Edition stamp beside the artwork.
The third pass is everything left, and it will be most of the box. Two groups come out separately as you go. Graded cards in sealed plastic are priced one at a time and never sampled with the rest. Anything already sitting in a sleeve or a toploader was put there by somebody who thought it mattered, so it joins the first pile until a sold listing says otherwise.
- Hits: holo or foil pattern, full art, gold or textured border, a 1999 to 2003 copyright line, a 1st Edition stamp
- Playable rares: a star rarity symbol in the bottom corner, no holo pattern, modern sets
- Bulk: circle and diamond rarity symbols, energy cards, and anything with a crease or a white corner
- Out of the count: graded slabs, sealed packs and boxes, and cards from a game that is not Pokémon
Each pile gets priced a different way
The three piles are not three sizes of one job. They are three different jobs. The hits carry the number, so each one is looked up on its own terms. The rares and the bulk are quoted per card, because a buyer will not pay you for the ten minutes it takes to price a $0.25 card and you should not spend them either.
Graded cards come off the total and get their own line. A PSA 9 and a PSA 10 of the same card are two prices, sometimes a long way apart, and the cert number on the label points at that exact copy. The same goes for sealed packs and boxes, which are priced as sealed product rather than as the cards that might be inside them.
Which odd-looking cards are errors
| Pile | What goes in it | How it gets priced | In the 3,000-card box |
|---|---|---|---|
| Hits | Holo, ultra rare, full art, 1999 to 2003, 1st Edition | One at a time, sold listings for the exact printing | 40 cards, $372 |
| Playable rares | Star rarity symbol, no holo, modern sets | A per-card rate quoted by a bulk buyer | 120 cards, $30 |
| Bulk | Commons, uncommons, energy, damaged cards | By the thousand or by weight | 2,840 cards, $28.40 |
| Graded slabs | Anything sealed in plastic with a grade on the label | One at a time, by grade and printing | Counted outside the box total |
Price the hits one at a time against sold listings
Take the first hit and read three things off it: the set symbol, the collector number in the bottom corner, and the language. Search on those, not on the Pokémon's name, because Charizard has been printed in dozens of sets and the prices have nothing to do with each other. Sold prices only. An asking price is what a seller hopes for, and the hope runs high on cards that have sat unsold for months.
Read the last ten sales and take the middle of them. One copy went for $60 because two bidders wanted it that night, six went between $22 and $27, and that card is a $25 card. Sort by date and treat anything older than about three months on a modern card as stale, because a reprint or a new set can move a price in a week.
Condition changes the hits and barely touches the bulk. A creased holo is a different card from a Near Mint holo, with its own much shorter list of sales, so price the copy in your hand against copies that look the same. Whitening along the back edges, surface scratches under a light, and a soft corner are the three things a buyer looks at first.
The arithmetic on a 3,000-card box
Here is the whole method on one box. Sorting 3,000 cards gives 40 hits, 120 playable non-holo rares, and 2,840 commons and uncommons. That is an evening's work. No prices yet.
Now price 20 of the 40 hits individually, from sold listings of the exact printing. Say the 20 total $186. Divide $186 by 20 and the average hit is $9.30. Apply that average to all 40 and the hits pile is 40 times $9.30, which is $372.
The 120 rares get a bulk quote of $0.25 a card, so 120 times $0.25 is $30. The 2,840 commons and uncommons go at a quoted cent a card, so 2,840 times $0.01 is $28.40. Add the three piles: $372 plus $30 plus $28.40 is $430.40. Call the box $430.
Then look at where the $430.40 came from. Forty cards out of 3,000 carry $372 of it, which is 86.4% of the total. The 2,840 commons carry $28.40, which is 6.6%. The 120 rares cover the remaining 7%. One pile decides the answer and the other two round it.
Every figure above is an input, and yours replace all of them. The 20 sold-listing lookups are the only real work in the method, and they are the reason the number is defensible instead of a guess.
How wrong $430 can be, and when to stop sampling
The estimate has one weak joint: the 20 hits you did not price. Suppose the second 20 are the weaker half and total $93 instead of $186. All 40 hits then come to $279, the box comes to $279 plus $30 plus $28.40, which is $337.40, or 22% below the $430 you quoted. That is the honest error bar.
It sits entirely in one pile, and widening the sample is the only thing that shrinks it. Price ten more hits and watch the running average. When the next ten cards move the total by less than a few percent, the average has settled and the rest of the pile can ride on it. On a box with only 40 hits, pricing all 40 takes about an hour and removes the error bar altogether.
Sampling the commons harder does nothing. If the entire 2,840 turned out to be worth double, the box would go from $430.40 to $458.80, a 6.6% change, while a bad read on 20 holos moved it by 22%. Spend the hour upstream, on the pile that decides the number.
A collection is a few cards and a lot of cardboard
This is the part most price guides skip. Value is not spread through a collection; it collects in a handful of cards and the rest is priced by weight. A 1999 Base Set holo and a 2019 energy card came out of packs that cost about the same, and today one of them carries the box while the other is counted by the thousand.
Two forces did that. Print runs grew enormously after the 1990s, so modern commons exist in numbers no collector base can absorb, and demand concentrates on a short list of cards: Charizard, first prints, 1st Edition stamps, and the chase card of whatever set is current. A card with no buyer has no price, whatever a price guide prints next to it.
So put the effort where the money is. The 40 hits are worth sleeves, toploaders and a box that closes; the 2,840 commons are worth a lid. Storage is the one input you still control after the sorting is done, and a summer in a loft is the fastest way to turn a hit into bulk.
Graded cards sit on their own line
A slab is priced by its grade as much as by the card, so pull graded copies out of the sampling and look each one up as a PSA 9, a PSA 10 or a CGC 9.5 of that printing. The cert number identifies the copy. Search sold listings for the card, the set and the grade together, because the gap between one grade point and the next is often larger than the gap between two sets.
The other question is whether to send anything from the box off to be graded. PSA's Standard service is $59.99 a card with a $1,000 declared value ceiling and a 90 to 100 business day turnaround, and its Value services are paused as of 2026-09-19. CGC's Bulk tier is $17 a card, but it needs 25 cards, caps declared value at $500, and runs 150 working days. Fees are not refunded for a low grade.
Against a box whose average hit is $9.30, that arithmetic is short. A $59.99 fee on a $9.30 card is not a decision. Grading is for the two or three cards at the top of the pile, if the box has any, and the way to settle it is to compare sold listings of the graded and the raw copy of that exact card.
Sources: PSA trading card grading services and fees, CGC Cards services and fees
Fees come off the $430 before you see it
What a collection is worth and what it pays are two numbers, and the gap is fees plus your evenings. eBay's final value fee in the trading card categories is 13.25% of the total amount of the sale up to $7,500 per item, plus a per-order fee of $0.30 on orders of $10.00 or less and $0.40 on orders above that. The percentage is charged on the item price, the shipping you charged the buyer, and the sales tax.
Run it on one hit. A $25 holo sold on eBay gives up 13.25% of $25, which is $3.31, plus the $0.40 order fee, so $21.29 is left before postage and packing. Do that 40 times and you have spent 40 evenings, which is the real reason a box that prices at $430 rarely returns $430.
Quote the collection at the market number, then say the net out loud as well. Anyone deciding what to do with a box needs both.
Where to sell Pokémon cards and what each venue keeps
Sources: eBay selling fees
Scan the hits and keep the total in one place
The slow part is the 40. CardLens matches the exact printing from a photo, with set, collector number, rarity and language, and shows a value range built on recent eBay sales of that printing, with the sales themselves visible one tap below the number. It estimates condition on the Near Mint to Damaged scale, and it reads PSA, BGS and CGC slab labels, so a graded copy is recorded with its grade rather than guessed at.
Saved cards go into a vault that keeps a running total, which is the job the half-finished spreadsheet was meant to do. Scan the hits, let the total build itself, and the estimate only has to lean on the two piles that get quoted by the card.
Questions collectors ask
How much is my Pokémon card collection worth?
Usually a lot less than the card count suggests, and almost all of it in a few dozen cards. The 3,000-card box worked through above came to about $430, of which $372 sat in 40 cards. Getting that number took an evening, not a fortnight, because only the 40 were priced one by one.
How do I value a Pokémon card collection?
Sort first, price second. Pull the holos, ultra rares and anything with a 1999 to 2003 copyright line, then price 20 of them from sold listings of the exact set and collector number and apply that average to the whole pile. Quote rares and commons per card.
Are bulk Pokémon cards worth anything?
Yes, by the thousand rather than by the card. Bulk sells by weight or by the box at a per-card rate the buyer quotes, and at a cent a card the 2,840 commons in the example came to $28.40. Ask for the rate before you spend a weekend sorting.
How do I value an inherited Pokémon card collection?
The same way, with one extra pass first. Look for graded slabs, sealed packs and boxes, and anything dated 1999 to 2003 before cards get moved or cleaned, since cleaning and pressing destroy value. Then sort into three piles and keep every graded card on its own line.
What is a Pokémon card collection spreadsheet for?
It holds the two things a total depends on: what each card is, meaning set, collector number and condition, and what comparable copies sold for, with the date you checked. Prices move, so an undated total is a guess. Any tool that stores those columns does the job.
Free on the App Store
Scan the hits and let the collection total add itself up.
CardLens matches each printing from a photo, shows a value range with the recent eBay sales behind it, and saves the card to a vault with a running total. Free on the App Store, with an optional premium subscription.