1994 — 2026 · a data study

The giant screen,
measured.

This is a study of every film released in IMAX from 1994 to 2026. The dataset has 1,406 films. It runs from small museum documentaries to the largest blockbusters ever made. The charts below show what got made, how it was filmed, and what made films succeed.

What is this dataset?
The films

Every movie shown in an IMAX cinema, from Wikipedia's IMAX release lists. This includes giant hits like Avatar, Oppenheimer and the Avengers films. It also includes small science-museum documentaries and films released only in China. In total: 1,406 films across 33 years.

The money

Box-office grosses come from Box Office Mojo. 900 films have a verified worldwide gross. Figures split into domestic (US and Canada) and international. All money is in US dollars, not adjusted for inflation.

The hardware

The count of IMAX screens worldwide comes from IMAX's own filings with the US Securities and Exchange Commission (SEC). This gives a screen count for most years from 2001 to 2026.

How to use it

Read top to bottom, or jump to any section. Each chart has a short "how to read this" note. Every IMAX term is defined in the glossary below. Hover over any point or bar to see the exact number.

films tracked
1,406
across 33 release years
peak year
2025
113 films — an all-time high
network 2001 → 2026
8.5×
225 → 1,865 systems
IMAX + Dolby premium
$372M vs $125M IMAX-only

Before the charts, the vocabulary. Every film here is "IMAX," but that word covers a four-rung ladder — from a movie physically shot on 70mm IMAX film down to an ordinary film digitally upscaled to fill the screen. The rung a film sits on shapes almost everything that follows, so it's worth two minutes up front. Ordered by how much real IMAX picture you actually get:

Full glossary — every IMAX term used on this page, defined tap to expand
I

Orientation

What we're looking at, and where it came from.

01

What counts as "IMAX," really

"IMAX" on a poster can mean four very different things — and the difference is the single biggest driver of everything that follows. It ranges from a film physically shot on 70mm IMAX cameras down to an ordinary film digitally upscaled to fill a big screen. Here's the ladder, with each tier's real median worldwide gross.

the core distinction DMR is not a different camera — it's a different screen. A DMR film was shot conventionally, then remastered and upscaled so it looks acceptable blown up to six storeys. The image is the same shape and resolution the director captured. A shot-on-IMAX film carries a negative physically larger than the projected picture, and opens the frame taller (to 1.43:1) so you literally see more of the scene. That extra picture — not just bigger — is what people are paying the premium for.
The "more picture" part, drawn to scale

Standard cinema is 2.39:1 — wide and short (dashed guide). IMAX 70mm opens the frame vertically: you don't get a wider image, you get a taller one. This is why an IMAX shot feels immersive rather than just large.

How much of each film is actually in full IMAX

Only ~20 films in history were shot on real IMAX film cameras — and even those switch to the tall 1.43:1 frame for only part of their runtime. This is the exact fraction in full-frame IMAX, per film, from Letterboxd's format archive. Hover for the gross.

Read: Dunkirk (75%) and Nolan's films lead; Oppenheimer is only 22% full-frame despite its reputation. "Shot in IMAX" almost never means the whole film — it means the director chose specific sequences to open up. The Odyssey (2026) is set to be the first-ever film shot entirely on IMAX film.

How these tiers are derived: the imax_65mm_minutes and imax_expanded_minutes columns tell us how a film was actually captured — minutes shot on real IMAX film vs minutes that merely open to a taller digital ratio. A film with 65mm-minutes > 0 (Dunkirk, Interstellar, Oppenheimer) is genuinely shot-on-IMAX; one with only expanded-minutes (Aquaman, Eternals) is certified-digital; everything else is DMR. The per-film 1.43:1 fractions above come from Letterboxd's format notes (letterboxd_imax_notes.csv).
a null finding worth stating Intuition says more IMAX footage should mean a bigger film. It doesn't. Across the 54 films with a measured IMAX-minutes value, the correlation between minutes-shot-in-IMAX and worldwide gross is 0.02 — essentially zero. How much of a film is in the tall frame is an artistic choice, not a commercial one. Oppenheimer (22% full-frame) out-grossed Dunkirk (75%); the format decision and the box office are independent.
02

The DMR pivot

For its first decade IMAX was purely IMAX 70mm — films shot on 15-perf 70mm for science museums. Then in 2002 DMR (Digital Media Remastering) let any Hollywood release be blown up to giant-screen. Watch the composition of the catalogue invert. New to these terms? The glossary above defines every one.

view
IMAX 70mm Filmed for IMAX IMAX DMR

Read: IMAX 70mm originals (amber) carry the format until ~2004, then flatten near zero. IMAX DMR conversions (gray) go from 0 to ~90 films a year. Toggle + network line to add the IMAX screen count from SEC filings. It uses the right axis, on a different scale. The screen build-out and the DMR film flood rise together.

what the data says In 2002 there were 225 IMAX systems worldwide and IMAX 70mm originals still outnumbered DMR conversions 16 to 6. By 2026 the network is 1,865 systems and the slate is almost entirely DMR. The screen build-out and the shift to conversions are the same story: a bigger network needs a constant supply of wide-release titles, which only DMR can provide.
II

What happened

The history: formats, the network build-out, and where screens landed.

03

Which premium format won

IMAX is one of several PLF (premium large format) brands. A film can play in more than one at the same time. Each format does something different. The table explains what. The chart then shows how their popularity changed over time.

Format
What it changes
How it works
3D
Adds depth to the image.
Two overlaid images and special glasses make the picture appear to have depth. Some films are shot in 3D; most are converted afterward.
IMAX with Laser
A brighter, sharper image.
Laser projectors replace older xenon lamps. They give a brighter picture and better contrast. The top "GT" version is the only digital system that shows the full, tall 1.43:1 frame.
ScreenX
A wider field of view.
The picture extends onto the left and right side walls for a 270-degree image. Made by the Korean company CJ 4DPLEX. It does not make the image taller, only wider.
4DX
Adds physical effects.
Seats move. Wind, water, scent and light effects match the action. Also from CJ 4DPLEX. It changes the seat, not the screen.
Dolby Cinema
Better contrast and sound.
IMAX's main rival, from Dolby. It uses Dolby Vision (very high contrast, over 1,000,000:1) and Dolby Atmos sound. It competes on picture quality and audio, not on a taller screen.
series
3DlaserScreenX4DX

Read: 3D (amber) peaks around 2016 at ~70% of the slate, then falls to the teens — the post-Avatar 3D bubble deflating in real data. 4DX and ScreenX climb as the newer premium tiers. Toggle series above; Y = % of that year's films, X = year.

what the data says 3D isn't a steady feature of IMAX — it's a wave. It crests in the mid-2010s and recedes. Meanwhile laser projection stays a thin premium sliver (10–18%), and 4DX/ScreenX pick up the "extra-format" role 3D vacated.
04

The maturing network

Two structural signals from the SEC filings that reveal a business shifting from novelty to saturation — and from museums to multiplexes.

Gross per screen is falling

Median film's worldwide gross divided by the number of IMAX screens live that year. A saturation signal.

Read: per-screen productivity peaked around 2010 then fell steeply. As the network exploded 8× and absorbed a long tail of smaller releases, each screen carries less median gross. Growth in screens outran growth in hits.

Museums squeezed out by multiplexes

Institutional (science-centre / museum) screens as a share of the total IMAX network, from the filings' 4-part breakdown.

Read: IMAX's origin was the museum dome, but institutional screens fell from 5.8% of the network (2018) to 3.1% (2026) as commercial multiplexes multiplied. The science-centre roots are being diluted — the same shift the DMR pivot showed in the film slate, now visible in the hardware.

05

Where the screens actually are

The slate is only half the business — the other half is hardware on the ground. Mapping the venue registries (imax_venues_worldwide.csv, dolby_venues_worldwide.csv) shows where IMAX built, where its premium-format rival Dolby Cinema went head-to-head, and how few screens can actually show the format at full height.

The global footprint

Every catalogued screen, placed on the map. The story is concentration — a few countries carry most of the network.

Read: the network is strikingly top-heavy. The United States (181), China (69), Japan (29) and Canada (28) hold most of the catalogued IMAX venues, trailing off into a long tail of single-screen countries. Dolby's footprint is smaller and even more US-weighted. Bubbles are placed at country centroids, not exact venues — city coordinates are not in the venue data, so this shows national totals, not street locations. The over-time shift toward China and international, described in IMAX's filings, is not chartable at country level from this snapshot.

IMAX vs Dolby Cinema, by country

Installed venues per country. The two premium formats fight hardest in the same handful of markets.

IMAX venues Dolby Cinema screens

Read: in the United States the two are near-parity (181 IMAX vs 178 Dolby) — the rivalry is fiercest in IMAX's home market. IMAX leads decisively in China, Canada and Japan; Dolby only ties or edges ahead in a few smaller European markets.

The network leans hard to the Americas

IMAX venues by world region. Nearly half sit in one region — the reach is broad but far from even.

Read: the Americas hold 45% of catalogued venues and Asia another 37% — together 82% of the network. Africa has just two. The map is a story of two continents.

Not all "IMAX" is the same IMAX

Venues by projection format-class. Only the IMAX 70mm / GT Laser (1.43:1) tier can open to the full-height frame that films like Oppenheimer and Dunkirk were shot for.

…and the rare frame is also the big frame

Median physical screen area by format-class, from the 342 venues with dimensions on file (imax_venues_worldwide.csv). The scarce 1.43:1 tier isn't just a taller ratio — it's a far larger picture.

Read: a median 1.43:1 screen is 447 m² (~18m tall) — 2.3× the area of a median 1.90:1 screen at 197 m² (~10m tall). The two dimensions of scarcity compound: the full-height frame is rare, and where it exists the wall itself is more than twice as big.

the scarcity finding Only 84 of 492 catalogued IMAX venues — 17% — can project the full 1.43:1 frame; the other 83% top out at the shorter 1.90:1 shape or the dome. And 57% of those 1.43:1-capable screens sit in the US and Canada. So when a film is shot in IMAX 70mm for the tallest ratio (Section 02), the overwhelming majority of the audience — everywhere outside North America especially — never sees it that way. The immersion ranking upstream is an aspiration most tickets can't deliver.
III

Why

What actually drives an IMAX film's reach.

06

Does capture type predict reach?

The revenue question, on 900 films with verified Box Office Mojo grosses. Each dot is one film — release year against worldwide gross, colored by capture type. Switch the y-axis to test whether the pattern survives once you remove the industry's overall growth.

y-axis
IMAX 70mm Filmed for IMAX IMAX DMR

How to read this: each dot is one film. The X axis is its release year. The Y axis is its worldwide gross on a log scale, so each gridline is 10× the one below. A log scale is used because grosses range from under $1M to nearly $3B. A normal scale would squash 90% of the films into a thin line at the bottom. Hover any dot to see the film. The ÷ year median view removes the effect of the industry growing over time.

what the data says Median worldwide gross: Filmed for IMAX $432M, IMAX DMR $166M, IMAX 70mm museum originals $13M. The museum-originals figure is misleading — those are old documentaries with tiny releases, not flops. The real signal is simpler. Filmed for IMAX blockbusters roughly double the IMAX DMR median. Much of that gap survives the year-median view. So it is a real format effect, not just recency.
Analysis honesty: capture type, release year, and network size all rise together — they're collinear. The ÷ year median view guards against reading recency as a format effect. Even so, Filmed for IMAX titles skew toward big-budget franchises, so part of the gap is budget, not the camera — only a multivariate model can fully separate those.
07

What actually drives reach

Capture type is one lever. Two others turn out to matter as much or more — and one of them cuts against the obvious read from the adoption chart. Both are computed on the 900 films with verified grosses.

Premium-format stacking

Median worldwide gross by how many extra premium formats (3D, ScreenX, 4DX, 70mm) a film shipped in.

Read: a clean monotonic ladder — $28M → $167M → $393M → $522M. More premium formats tracks with far higher gross. It's largely selection (studios reserve the full format stack for their biggest titles), but that makes format-count a strong signal of a tentpole.

The 3D premium didn't die

Median gross, 3D vs non-3D films, across three eras. The adoption chart showed 3D releases falling — but look at the gross.

3Dnon-3D

Read: fewer films get 3D now. But the ones that do are the big ones. The 3D-vs-non gap widens from about 2× to nearly ($432M vs $46M in 2020–26). 3D became a marker of a blockbuster, not a fading gimmick.

the counterintuitive one Section 03's adoption curve made 3D look like a dying fad — its share of releases collapsed after 2016. But share of releases and share of revenue are different questions. 3D stopped being applied to mid-tier films and concentrated onto tentpoles, so per-film it looks more premium than ever. Same data, opposite conclusion depending on which denominator you pick.
Median gross per film, by era

The rise — and recent fall — of the typical IMAX film's reach.

Read: the median climbs from $13M (pre-2005) to ~$300M (2010–14), then crashes to $84M in 2020–26. That fall isn't IMAX weakening — it's volume: 378 films in the latest bucket vs 138 in 2010–14. The network now absorbs a long tail of small releases alongside the tentpoles, dragging the median down even as the top end grows.

IMAX vs Dolby — the co-release premium

IMAX isn't the only premium screen. 135 films also got a Dolby Cinema release. Films that land both formats gross far more than IMAX-only.

IMAX onlyIMAX + Dolby

Read: $372M vs $125M — a 3× gap. Studios save the full premium stack — IMAX and Dolby together — for their surest hits. So a dual-format release is one of the strongest signs of a major title.

Dual-format release is rising fast

Share of each year's IMAX slate that also gets a Dolby Cinema release, since Dolby Cinema scaled up.

Read: from effectively 0% before 2017 to ~30% by 2023–26. The two formats increasingly co-exist on the same tentpoles rather than competing — the premium tier is consolidating, not fragmenting.

On the comparison: the IMAX-vs-Dolby gross gap is selection, not causation — Dolby doesn't make a film gross more; studios grant both formats to films already expected to be huge. It's a reliable signal of a tentpole, not a lever you can pull. The Dolby list (179 titles) comes from Wikipedia's Dolby Cinema category; the Wayback reconstruction of Dolby's own site returned empty, so this is the film-list angle only, not a screen-count time series.
Best available projection tier vs gross

Median worldwide gross by the highest projection format a film was released in — from ordinary digital xenon up to the flagship GT dual-laser system.

Read: a clean 4× climb — $124M (digital xenon) to $521M (GT dual-laser). The premium-projection venues get the premium films. Note the small counts on the top tiers (n=20 each): these are the handful of flagship releases studios push into the best houses.

08

Franchise economics

The strongest single predictor of an IMAX film's reach isn't the camera or the projector — it's whether it belongs to a franchise. Detected from title patterns across 900 films with grosses.

Franchise vs standalone

Median worldwide gross, franchise entries vs original standalone films.

Read: $709M vs $108M — a 6.5× gap, wider than any format effect in this dashboard. IMAX's commercial model runs on franchise tentpoles; standalone films are the long tail.

The biggest franchises on IMAX

Median gross per film, franchises with 3+ IMAX releases. Bar length = median; hover for total franchise haul.

Marvel/MCUother franchises

Read: Avatar tops per-film median, but Marvel is the volume engine — 41 IMAX films totalling $35.5B, more than the next three franchises combined.

Method: franchises are matched by keywords in the title (e.g. any film containing "Avengers", "Spider-Man" etc. maps to Marvel/MCU). This under-counts films with non-obvious titles and can't see franchises not in the keyword list, so treat the standalone bucket as "franchise not detected" rather than "definitely original." Even so, the gap is large enough that reasonable mis-classification wouldn't erase it.
09

Relationships that actually hold

Rankings show what's bigger; these scatters show what predicts what. Each dot is a real data point. Each axis is a real quantity. The dashed line is the fitted trend. Its correlation r runs from 0 (no relationship) to 1 (perfect).

Infrastructure drives supply

Each dot is one year. X = IMAX screens installed worldwide, Y = films released that year.

one yearfitted trend (r = 0.93)

Read: a near-straight line, r = 0.93. Every ~100 new screens brings roughly 5 more films a year. The film supply is almost mechanically tied to how many screens exist to fill — the strongest relationship in the dataset.

The opening weekend forecasts the run

Each dot is a film. X = domestic opening weekend, Y = final worldwide gross (both $M, log-scaled).

one film (n = 717)

Read: a tight upward band, r = 0.82. A film's first three days strongly predict its entire global haul — which is exactly why the industry obsesses over opening weekend. Outliers above the line are the rare films with "legs" that over-performed their start.

Franchises get the better formats

Capture-type mix within franchise films vs standalone films. Stacked to 100% so you compare composition, not volume.

Filmed for IMAXIMAX DMRIMAX 70mm

Read: franchise films are 4× more likely to use Filmed for IMAX capture (20% vs 5%) — studios invest in the premium format precisely where the audience is guaranteed. Standalone films lean on the cheaper IMAX DMR path or are museum originals. Format choice follows commercial confidence.

IV

So what

The money, and the shape of the business.

10

Where the money comes from

Box Office Mojo splits gross into domestic (US / Canada) and international — not per-country, but enough to see IMAX's center of gravity shift. As the network built out overseas, the international share of gross climbed and stayed high.

view

Read: international share of every year's total worldwide gross for films with a reported split. It roughly doubles from 2000 to the mid-2010s, then plateaus around two-thirds. vs network size overlays the installed-system count to show they grew together.

what the data says In 2000, international was 28% of IMAX-film worldwide gross; by the 2010s it settled near 65%. IMAX became an internationally-driven business — the overseas screen build-out (especially China) is where the growth went. The china_only flag marks 177 films that never got a domestic IMAX release at all.
Data limit: this is a two-bucket split (domestic vs international), not per-country. A true country-level map isn't supported by these sources — Box Office Mojo's per-territory tables weren't captured, and the only finer geographic signal in the pipeline is the binary china_only flag. Claiming per-country revenue here would mean inventing it.
11

The economics behind the frame

Joining production budgets and original language from tmdb_enrichment.csv exposes two things the release data alone can't: which capture format returns the most per dollar spent, and how the slate stopped being a Hollywood-export box and went genuinely multilingual.

Return on budget, by capture type

Median worldwide gross ÷ production budget, for the 660 films with a real (>$1M) budget on record. Higher = more reach per dollar.

Read: the prestige 65mm-film titles (amber) aren't just an artistic flex — at a 3.5× median return they're the most budget-efficient tier, ahead of DMR (2.85×) and certified digital (2.66×). Native museum originals trail at 1.86×, on tiny budgets. Native n=7 — read as indicative, not firm.

The slate went multilingual

Films per year by original language (those with a verified gross). Watch the English share erode after 2019.

English Chinese Japanese other

Read: non-English titles were under 10% of the slate through 2010–14 and are ~48% by 2025–26 — Chinese, Japanese and a fast-growing "other" bucket (Hindi, Korean, Telugu, French).

what the data says The two findings connect. IMAX's international share of gross (Section 10) rose because the network globalized — and here's the supply-side proof: the company now programs a slate that is nearly half non-English, feeding screens in China, Japan and India that a Hollywood-only pipeline could never fill. Meanwhile the tier IMAX markets hardest — true 65mm film — quietly returns the most per dollar, which is why the expensive "shot-for-IMAX" prestige play keeps paying off even as the network saturates.
12

Release rhythm & concentration

When IMAX films come out, and how unequally the money is shared. Release dates parsed from 1,042 films; grosses from 900.

Seasonality — count and median gross by month

Bars show how many films release each month; the line shows their median gross. The tentpole calendar is unmistakable.

films releasedmedian gross

Read: a clear two-peak tentpole rhythm — summer (May–July, median ~$370M) and the November holiday launch ($261M). January is the dead zone ($52M). Studios save their IMAX-worthy films for the windows audiences show up.

The box-office pyramid

How the 900 films distribute across gross bands.

Read: 52 billion-dollar films sit atop a broad base — 233 films earned under $25M. IMAX is a hit-driven business: a few giants, a long tail.

Revenue concentration

Share of all IMAX-film worldwide gross earned by the top 10% of titles.

top 10% of filmsother 90%

Read: the top 10% of films (about 90 titles) earn 40% of all the box office across $233B total. The blockbusters carry the network.

13

What IMAX itself reports

The sections above measure the slate. This one measures the company. It reads IMAX's own numbers straight from its SEC filings (imax_economics_from_edgar.csv, 103 filings, 2002–2026) — the box office its network actually captured, quarter by quarter.

Quarterly gross box office, as reported to the SEC

Each point is a figure IMAX disclosed in a quarterly or annual filing. The gap in 2020–2021 is the pandemic, when theatres closed and filings carried no comparable figure.

Read: IMAX box office grew from about $17M a quarter in 2002 to a $200–370M range through the 2010s. The 2020–2021 gap marks the pandemic shutdown. Recovery has been strong — the two highest points in the series, $347M (Q3 2023) and $368M (Q3 2025), are both post-pandemic.

the billion-dollar line In its 2015 full-year filing, IMAX reported that its global box office passed $1 billion for the first time — a 31% jump on the prior year. The quarterly series shows how it got there: a steady climb through the early 2010s as the network scaled and the blockbuster slate matured. The recent record quarters show the business is not just back to its pre-pandemic level but above it.
V

Your turn

The producer's verdict — why a film green-lights IMAX.

14

Why a producer green-lights IMAX

Every section above measures what happened. This one asks why — the decision a producer actually makes. It is a cost-versus-payoff calculation, and it explains why IMAX 70mm capture stays rare and depends heavily on the specific film. The framing draws on trade reporting from The Oklahoman (2017) and a 2025 large-format cinematography primer, checked against this dataset.

The commitment funnel

How far a film goes into IMAX narrows sharply at each step. Most "IMAX films" never touch an IMAX camera.

Read: of 1,406 IMAX releases, only 408 (29%) were shot with large-format cameras — the rest are DMR post-conversions. Only 38 films (3%) commit to the full 1.43 frame. This mirrors the trade estimate that fewer than 15% of a given year's ~30 IMAX releases use IMAX's own cameras.

Camera capture is a big-budget privilege

Median production budget, camera-shot films vs DMR-only conversions. The gear only makes sense at the top of the market.

Read: camera-shot films carry a median budget of $165M — nearly double the $85M of DMR-only films. 67% of camera-shot films cost $100M or more, matching the reporting that studios rarely go full IMAX below a ~$100M budget.

The producer's ledger

The decision weighs concrete costs and logistics against a real commercial upside. Both sides are drawn from the trade sources above.

Reasons not to (why it stays rare)
  • Camera cost. An IMAX camera rents for roughly $40–50k per week; each ~3-minute film magazine runs about $3,000, and scanning adds ~$20k per hour of footage.
  • Scarce, difficult gear. Only about ten 15/70mm film cameras exist worldwide. They are heavy, run loud enough to complicate live sound, and hold only ~3 minutes of film per load.
  • Screen mismatch. Only ~100 theatres worldwide have the 1.43 screen. Shooting 1.43 means giving up roughly a quarter of the frame in most houses (see Section 12).
  • Rigid workflow. Native capture generates enormous data and demands specialised lighting, lenses, and a locked delivery spec — slower and costlier than a normal shoot.
Reasons to (why the top of the market does it)
  • Box-office share. IMAX runs about 10% of opening-weekend grosses for big blockbusters, and films built with IMAX "DNA" can pull a much higher share.
  • Longer legs. Films that use IMAX technology tend to hold their IMAX screens for more weeks than a typical release.
  • A reason to leave the couch. As home screens improve, native capture is marketed as the experience audiences cannot get at home.
  • Efficient at the top. In this dataset the camera-shot / 65mm tier returns the most per budget dollar (Section 11) — the spend pays back for the films that can carry it.
The payoff, made concrete. When IMAX reports its share of a film's box office, the "punches above its weight" claim becomes specific. These are the disclosed figures in the data.
the answer A producer green-lights IMAX 70mm capture when the film can carry the cost and exploit the payoff — a big-budget, spectacle-driven title where immersion is central and a ~10%+ opening-weekend IMAX share is worth chasing. That is why it is a creative-and-commercial choice, not a technical default. It also explains the shape of the whole dataset: 71% DMR conversions, a thin 3% using the full frame, and camera capture clustered among the most expensive films. IMAX is not used commonly because, for most films, the ledger does not balance.