EX Era Pokémon Cards Value: All 1,694 Cards

Between 2003 and 2007 the English game printed sixteen expansions under the EX banner, from [Ruby & Sapphire](/research/set-screen/ruby%20%26%20sapphire) to Power Keepers. We hold 1,694 English cards across them. On September 23, 2026, 1,688 of those carried an asking price on our daily tape, summing to $87,845.77. The distribution is the finding. Sixteen cards asking $1,000 or more carry 32.6% of that total, and the 1,040 cards asking under $5 carry 1.8% between them. The median EX-era card asks $2.48. Fourteen of the sixteen four-figure cards are Gold Stars, and the six cards in the whole era carrying no reading on the census day are Gold Stars too.

Every card in every set is here. Nothing is filtered on price, rarity or registry depth, and the six unpriced cards are named later rather than dropped quietly. Summed ask is the sum of the day-20719 tape over the priced products in the set, which is not the same number as the "Full Set" figure on a set page: that comes from a separate recorded value series, and on Power Keepers the two read $4,794 and $4,352.03. Only the tape is used below.

The ordering is not the order the sets were printed in, and it is not size either. [Ruby & Sapphire](/research/set-screen/ruby%20%26%20sapphire) has more cards than [Legend Maker](/research/set-screen/legend%20maker) and a third of its value. What separates them is the top of the checklist: Legend Maker holds three Gold Stars and Ruby & Sapphire holds none. Across the sixteen sets, a Gold Star on the checklist tracks the total more closely than card count does. Nine of the sixteen printed one, and those nine hold ranks 1 to 8 and rank 10 of this table: [FireRed & LeafGreen](/research/set-screen/firered%20%26%20leafgreen) at ninth is the only set without a Gold Star that outranks a set with one. Seven never printed a Gold Star, and four of those seven are the four sets at the bottom of the table.

A census over a whole era is worth publishing mostly because the shape is not what a top-ten list implies. 1,040 of the 1,688 priced cards, 61.6% of them, ask under $5, and they add up to $1,621 of the era's $87,845.77. At the other end, sixteen cards hold $28,661.

The nine bands hold 1,688 cards and $87,845 between them, which is the whole priced population. That is the point of printing the distribution rather than a leaderboard: the ten dearest cards are 24.5% of the era, the fifty dearest are 53.1%, and the hundred dearest are 68.3%. Everything below that is a $2 card. How much of a set is bulk is measured across the modern catalogue in Which Pokémon Sets Are Mostly Bulk, and the EX era sits at the bulk-heavy end of it.

Our rarity column is useless for this era. It calls Torchic ☆ at $4,500 and a 50-cent Hoppip both "Rare", because the EX-era print runs predate every modern rarity tier. So the tier below is read off the printed card name instead, and the rule is published here so it can be checked: a name carrying the glyph ☆ or ★, or ending in the word Star, is a Gold Star; a name carrying ex as its own word is an ex, and an ex also carrying δ is an ex δ; a name carrying δ without ex is a Delta Species card; everything else is ordinary. Matching on the glyph alone misses five cards, because the catalogue spells Groudon Star, Kyogre Star and Metagross Star in Delta Species and Mewtwo Star and Pikachu Star in Holon Phantoms with the word rather than the symbol. There are 25 Gold Stars in the sixteen sets, not the 20 a glyph search returns.

The five tiers hold all 1,694 cards. Two columns move in opposite directions down the table and neither is a coincidence: the median ask falls from $1,333.56 to $1.86, and the median gem rate rises from 3.8% to 12.6%. That inverse relationship is the whole of our companion study EX Era PSA 10 Gem Rates, which measures it over the 519 cards with enough recorded submissions to support the figure. The δ mechanic itself, 191 cards scattered across eight sets, was measured separately in A 191-Card Set Nobody Printed; the 179 δ cards in these sixteen sets are the EX-era subset of it, 176 of them in the two δ tiers above and three more filed here as Gold Stars.

Ranks 1 to 16 are printed contiguously and are the whole population of EX-era cards at $1,000 or more; the seventeenth is Metagross Star at $951.99. Eight of the sixteen carry no PSA 10 sold median, meaning fewer than three verified sales are on record for the grade, which is the same pattern our whole-catalogue study 14,466 Cards With a Price and No Sale finds at the expensive end of vintage. We print the asks here and nothing else about the Gold Star tier: the ladder, the graded coverage and the question of which one is the cheapest route into the run belong to the cheapest way to own a Gold Star, which covers all 27 of them including the two in POP Series 5 that are outside this era and are measured in POP Series Pokémon Cards.

On the same day the English catalogue holds 21,416 priced products summing to $439,130, with a median card of 89 cents. The EX era is 7.9% of those products and 20.0% of that value. Twenty-five of the hundred dearest English cards come from these sixteen sets, and thirteen more come from the three e-Card sets printed immediately before them, so 38 of the English top hundred were printed between 2002 and 2007. Torchic ☆ is tied with Lugia 149 from Aquapolis for the dearest English card we price, both at $4,500.

Three neighbouring populations are deliberately outside this count and are measured separately, because folding any of them in would change every number above. The Japanese printings of these same cards are a different catalogue with different prices, covered in Vintage Japanese Pokémon Card Prices. The English Black Star Promo bins that ran alongside the EX sets are not expansions and are counted in Are Black Star Promos Worth Anything; one of them, a Nintendo Black Star Promos Mudkip, sits in the English top twenty at $1,500 and is not in this study. And the eras on either side have their own censuses: the Neo sets before, Diamond and Pearl and Sun and Moon after, plus the POP Series decks that ran concurrently with the EX run and hold the two Gold Stars this study does not.

Two things this does not mean. It does not mean the era outperformed, because nothing here is a return: the whole EX-era tape is 25 days old and every card's weekly series begins on August 30, 2026, so there is no window to measure. And it does not mean the era is liquid. Over the sixteen days from September 7 to September 23, both of them complete tape days, the median EX-era card moved 0.00% and 742 of the 1,688 did not change by a cent. A price that has not moved in sixteen days is a listing nobody has taken.

Three observations would break the central claim, which is the concentration and not the total. First, a sold-price census: every figure above is an asking price, and if verified sales on the sixteen four-figure cards came back at a fraction of their asks while the $2 tail sold at its ask, the 32.6% would shrink toward the tail. That test is partly available already, because the eight of the sixteen that carry a recorded PSA 10 sold median all sell in the slab for a multiple of the raw ask, from Lugia ex at 6.6x to Pikachu Star at 52.3x, which points the other way. Second, a second printing tier we have mis-scoped: if a meaningful number of the 1,365 cards in the bottom tier are being quoted at the wrong printing, the tail is worth more than $16,995 and the concentration is overstated. Third, a coverage gap: if the era holds cards we do not have in the catalogue at all, and they are chase cards rather than commons, every share here is too high. The check available without a physical checklist is the printed set total, and on all sixteen sets our card count equals it or exceeds it by one to four, which is the shape a secret slot leaves, and never falls short: [Unseen Forces](/research/set-screen/unseen%20forces) prints 115 and we hold 117, [FireRed & LeafGreen](/research/set-screen/firered%20%26%20leafgreen) prints 112 and we hold 116.

It is one day. Raw prices are asks from TCGplayer's market figure for each product's canonical printing, not sold prices, and on a card with a single listing the ask IS the quoted market to the cent, which our study Five Cards in Six Trade on a Single Listing measures across the catalogue. Graded medians are the opposite instrument, sold prices over verified sales, and the two are never mixed in one figure here. All sixteen sets carry a recorded release date, from Ruby & Sapphire on June 18, 2003 to Power Keepers on February 1, 2007, but our daily tape begins on August 30, 2026, so there is no window between any launch and any reading used here: nothing in this piece is windowed to a launch or an anniversary, and no age comparison is made. The per-card weekly series for these cards starts on August 30, 2026, which forecloses every return, drawdown and since-launch figure that our other set studies carry. Set membership is the catalogue's `set_name`, so a card that a collector files under a different parent set is counted where we file it. The six unpriced cards are excluded from every price figure and named above rather than zeroed. Nothing here is investment advice.

Population: `cards` rows with `lang = 'en'` in the sixteen set names Ruby & Sapphire, Sandstorm, Dragon, Team Magma vs Team Aqua, Hidden Legends, FireRed & LeafGreen, Team Rocket Returns, Deoxys, Emerald, Unseen Forces, Delta Species, Legend Maker, Holon Phantoms, Crystal Guardians, Dragon Frontiers and Power Keepers: 1,694 rows. Deduplicated on `tcg_id`, which collapsed zero rows here and 1,212 across the whole catalogue, so the count is products rather than rows. Prices are `price_history.price` on day 20719, which is 2026-09-23 and the largest of the tape's 25 recorded days at 52,378 rows; days 20704 to 20713 carry 35,700 to 50,400 rows and are partial dumps, and none of them is used. 1,688 of the 1,694 carry a row on that day; the six that do not are named in the article and excluded from every price figure rather than treated as zero. `price_history` records each product's CANONICAL printing under the shared price rule in `packages/core`, deliberately not the dump's flagship figure, so a multi-printing product is quoted at the printing the market trades rather than at its dearest row. Medians take the upper middle value on even samples throughout. The name tier is a rule over `cards.name` after no folding is needed, because the catalogue spells these names consistently: `[☆★]` anywhere or a trailing word `Star` is a Gold Star (25 cards, of which 20 carry a glyph and 5 spell the word); a standalone word `ex` plus `δ` is ex δ (11); a standalone word `ex` alone is ex (128); `δ` alone is Delta Species (165); the remaining 1,365 are ordinary. The tiers partition the 1,694 exactly. PSA 10 medians come from `graded_snapshots`, latest row per card and grade across both the catalogue-id and TCGplayer-id keyspaces with the fresher row taken, and are SOLD medians over verified sales rather than asks. A graded row is dropped when the fresher of the two keyspaces still predates the tape's first day, 20695; across the 2,227 products of this ground that rule removes seven PSA 9 and PSA 10 medians on five cards, three of them PSA 10 and none of them printed here, and every one of the seven sits in the catalogue-id keyspace. They are named in the Team Rocket Returns and gem-rate companion studies. Gem rates come from `population_snapshots`, latest day per key across both keyspaces, days 20704 to 20719 in this ground, and pass `gemRateIsTrustworthy()` from `@cardapp/core/server` with the recorded PSA 10 sold count supplied; the tier table's gem column is the median over the cards in that tier carrying 100 or more recorded submissions. English catalogue context is the same day-20719 tape over all `lang = 'en'` rows deduped on `tcg_id`: 21,416 priced products, $439,130. The sixteen-day change compares day 20703 to day 20719, both complete days; fifteen of the tape's 25 days are complete and the ten between 20704 and 20713 are not. Release dates come from `api_cache` key `report:spread:setsmeta`, which spells these sets `EX <name>` and `Expedition`, so a match on the catalogue spelling alone finds none of them. Set and card links were fetched from production on 2026-09-23 and their page titles read back; `/set/dragon` returns 404 and `/set/expedition base set` returns 404, so neither is linked. Scripts: apps/web/scripts/vex-build.mts and vex-figures.mts; every printed figure with its query is in apps/web/.scratch-vintage-ex/figures.json. Nothing here is investment advice.