July 20, 2026 · 9 min read · las vegas · casino economics · blackjack

Where Las Vegas Strip money actually comes from

The Las Vegas Strip is understood as a gambling business the way Detroit is understood as a car business: accurately for a while, then out of habit. The regulatory filings tell a different story, and they have been telling it for a quarter of a century. The fountains, the residencies and the sphere-shaped screens are not paid for at the tables. They are paid for at the front desk.

This is not an argument that requires interpretation. Nevada is unusual in that its casinos file audited departmental income statements with the state, and the Gaming Control Board publishes them. So the question of where Strip money comes from has a filed answer, broken out by department, updated every year. Here is what the most recent filings say, and what they imply about the small corner of the floor where skill still changes the outcome.

The number everyone quotes

Las Vegas Strip casinos won $8,815,212,000 from gamblers in calendar 2025, according to the Gaming Control Board's December 2025 revenue report. That is the headline figure, the one that moves markets and gets reprinted every month. Against 2024 it was up 0.03% — not a decline, not a boom, essentially the same year twice. Win amounts in these reports are stated in thousands and cover the Strip reporting area only, which is a subset of Clark County.

It is a real number and it measures a real thing: money that patrons lost and casinos kept. What it does not measure is the size of the business. Gaming win is a single line on a much longer income statement, and on the Strip it has not been the biggest line for a long time. Treating it as a proxy for how Las Vegas is doing is like tracking an airline by its baggage fees — the figure is precise, published, and answers a smaller question than people think.

The mistake matters because the two figures move differently. Gaming win is volatile, driven by a handful of very large players and by whether the dice cooperated in a given quarter. Room and food revenue is driven by occupancy, rates and convention calendars. When you read that Las Vegas had a good year or a bad one based on gaming win alone, you are reading about roughly a third of the business.

Gaming is about a quarter of the Strip's revenue

The Nevada Gaming Abstract is the document that settles this. For the fiscal year ended June 30, 2025, the 51 Strip casinos grossing $1 million or more reported total revenue of $21,077,838,684. Gaming accounted for $5,509,848,670 of it, or 26.1%. Rooms accounted for $7,064,968,548, or 33.5%. Rooms out-earned gambling by $1,555,119,878.

The full departmental split for fiscal 2025 on the Strip runs as follows.

Food and beverage together came to $5,573,194,801, which also exceeds the gaming line. The Strip, on this accounting, is a hotel and restaurant business with a casino attached — and one of the largest such businesses on earth. Note the fiscal year runs July to June, so these figures do not line up with the calendar-2025 gaming win above; they are different periods and different filings, and the rest of this post keeps them separate.

Rooms: $7,064,968,548 (33.5% of total revenue)
Gaming: $5,509,848,670 (26.1%)
Food: $3,993,791,473 (18.9%)
Other: $2,929,826,665 (13.9%)
Beverage: $1,579,403,328 (7.5%)

Why 26% and 35% are both correct

That 26.1% is lower than most published figures, and the reason is comps. The Abstract's casino department detail shows gross casino revenue of $8,444,797,191 for the Strip in fiscal 2025 — table games including keno and bingo at $2,990,691,577, slots at $5,102,997,347, card games at $173,455,229, race book at $11,062,023 and sports book at $166,591,015. From that it subtracts $2,934,948,521 of contra revenue, the accounting name for complimentary expense. What remains is the $5,509,848,670 gaming line. Across all departments, the Abstract puts complimentary expense at 38.89% of gaming revenue measured before that deduction.

The UNLV Center for Gaming Research handles this differently. To keep its long historical series consistent across a 2019 accounting change, it adds comps back into gaming revenue, on both sides of the ratio. Applying that method to the fiscal-2025 Strip figures — my own arithmetic, not a UNLV publication — gives a gaming share of about 35.2%. UNLV's published figure for fiscal 2023 was 35.04%, so on a consistent basis the Strip's gaming share has flattened out at roughly a third rather than continuing to slide.

The comps also cut the other way, and this is the part that undercuts a tidy narrative. Of the Strip's fiscal-2025 room revenue, $1,716,006,653 was complimentary room sales — rooms given away to gamblers, still booked as room revenue. Complimentary beverage sales were $574,656,754, which is 36.4% of the entire beverage department. Complimentary food was $586,505,707, or 14.7% of food. The hotel is out-earning the casino partly by billing the casino.

Forty years of drift

UNLV's departmental series starts in fiscal 1984, and the shape of it is unambiguous. In 1984, gaming was 58.63% of Strip revenue and rooms were 16.10%. By fiscal 2023 gaming was 35.04% and rooms were 29.38%. Food rose from 11.41% to 15.86% over the same span. UNLV's own summary of the Strip describes gaming win 'dropping from 59% of total revenue to 35%.'

The crossover year in that series is fiscal 1999, when gaming fell to 48.08% and the non-gaming departments together out-earned gambling for the first time in the data. That is also, not coincidentally, the era of the Bellagio, the Venetian and Mandalay Bay — properties built on the theory that a resort could charge resort prices. Every year from 1984 through 1998 sits at or above 50.25%; the descent from 1994 onward is steady.

It is not a straight line after 1999, and anyone presenting it as one is selling something. Fiscal 2021 bounced back to 48.03%, because covid gutted room and food revenue far more thoroughly than it gutted gambling. The denominators collapsed and gaming's share rose without gaming doing anything. Statewide the same drift is milder: Nevada as a whole went from 61.95% gaming in fiscal 1984 to 43.58% in fiscal 2023. The Strip is the extreme case, not the typical one.

Inside the gaming half, the machines win

Now zoom into the gambling itself. Of the Strip's $8,815,212,000 in calendar-2025 win, slot machines produced $4,950,173,000 and the category the state calls Table, Counter and Card Games produced $3,865,039,000 — 56.2% against 43.8%, from 35,232 slot units against 2,616 table units. On the Abstract's fiscal-year basis the gap is wider still: slots at $5,102,997,347 against table games at $2,990,691,577, or 1.71 to 1.

The table-game figure is also flattered by what is bundled into it. The Strip's Table, Counter and Card Games category includes sports pool at $271,324,000, card games at $178,023,000, race book at $8,826,000 and keno at $2,122,000 for calendar 2025. Those four are 11.9% of the category and 5.2% of all Strip gaming win. Strip out everything that is not a dealt table game and you are left with $3,404,744,000. Nevada statewide is more machine-dominated still, with slots taking 67.8% of the $15,798,466,000 won in calendar 2025.

One statistic runs the other way, and it deserves its moment. Per square foot of floor space in fiscal 2025, Strip table games generated $4,870 against $2,366 for slots. Tables earn more from every foot they occupy; there are simply far fewer feet of them — an average of 18,063 square feet of table space per casino against 59,906 for slots. The machines do not win on productivity. They win on being everywhere, all the time, needing no dealer and no minimum bet.

Slot machines, calendar 2025: $4,950,173,000 on 35,232 units
Table, Counter and Card Games: $3,865,039,000 on 2,616 units
Of which sports pool: $271,324,000; card games: $178,023,000
Of which race book: $8,826,000; keno: $2,122,000
Actual dealt table games, net of the above: $3,404,744,000

Baccarat, not blackjack

Within the tables, the story is not the one most players carry around. The Strip's biggest table game in calendar 2025 was baccarat, which won $1,409,598,000. Blackjack — 'Twenty One' in the filings — won $897,035,000, down 12.57% year over year. Baccarat earned about 57% more than blackjack, and it did so from an average of 392 units against blackjack's 1,034. That is not a close contest; it is a different order of business being conducted at a quarter of the tables.

This is a reversal, and it is dateable. In 1985, blackjack was 77.22% of Strip table units and 50.25% of Strip table win, while baccarat was 13.60% of the win. Baccarat passed blackjack in 2009, taking 37.91% of Strip table win against blackjack's 28.28%. By 2024 the split was 40.45% baccarat to 30.43% blackjack, and blackjack's share of units had fallen to 48.48%.

Craps shows the same effect from a different angle. Its share of Strip table win fell from 28.04% in 1985 to 9.82% in 2024 — but in dollars, Strip craps won $332,309,000 in calendar 2025, up 0.32% on the year. Craps did not collapse. The pie around it grew several times over, and the game did not grow with it. Share and dollars tell opposite stories, which is why the choice of which to quote is usually an argument in disguise.

Hold is not the house edge, and why that matters

The revenue reports list a win percentage next to each game, and it is the most misread column in Nevada gaming. Strip blackjack shows 12.21% for calendar 2025 and Strip slots show 7.83%. Read casually, that says blackjack is the worse bet. It says nothing of the kind, and the Control Board's own definitions explain why: for slot devices the figure 'represents the reported win amount divided by the total dollar amount played by patrons,' while for games it 'provides a ratio which has been adjusted for effects of credit play.' Different denominators. Not comparable, and neither one is a house edge.

The house edge is a separate measurement — the ratio of the average loss to the initial bet. On that measure the Wizard of Odds puts blackjack at 0.28% under a liberal Las Vegas rule set played with perfect basic strategy: dealer stands on soft 17, double on any two cards, double after split, resplit aces, late surrender. Baccarat's banker bet is 1.06%, the craps pass and come bets are 1.41%, double-zero roulette is 5.26%, and slot machines run roughly 2% to 15%. Change the blackjack rules and the 0.28% moves; the point is not that any real table is free, but that the ranking by house edge looks nothing like the ranking by reported win percentage.

That gap is the whole reason skill survives on a casino floor at all. It only exists in games where your decisions change the expectation — and those games are a shrinking slice of a shrinking slice. Statewide, blackjack tables fell from 3,682 in 2000 to 1,947 in 2025, a 47.1% decline, while blackjack win rose just 2.09% in nominal dollars over the same 25 years. Total table units fell 21.1% in that period, so blackjack lost floor space both outright and relative to the other games. The tables that remain each work far harder: win per blackjack unit went from $318,958 to $615,815.

Put the two halves together and you get the number that frames everything above. Blackjack is about 10.2% of Strip gaming win, and gaming is about 35% of Strip revenue on a comps-added-back basis, so blackjack is roughly 3.6% of all Las Vegas Strip revenue. That is my own calculation, mixing calendar-2025 gaming data with a fiscal-2025 revenue mix, and it deserves one significant figure rather than three. Call it a bit over three percent. The most famous game in the building is a rounding error in the business — which is exactly why the casino has so little reason to make it easy, and exactly why the players who bother to learn it properly are still the ones the surveillance room watches.

Blackjack is three percent of the Strip and the only meaningful percentage you control. Learn the count on a trainer that keeps score, then decide what a table is worth before you sit down.

Train the count →

Share this post — the link unfurls with its own card.