21+3 is a separate side bet attached to blackjack. It combines the player’s first two cards with the dealer’s upcard and evaluates those three cards as a three-card poker-style hand.
Use this guide for the full 21+3 wager, not just the term
This page owns the complete 21+3 betting decision: which three cards are used, how qualifying hands are ranked, how paytables change expected value, what an eight-deck probability model looks like, and how the extra wager changes total action at a blackjack table. If you only need the one-paragraph meaning of the term, use the 21+3 glossary entry. Keeping the definition and the full pricing/procedure guide separate makes it easier to distinguish “what is this bet?” from “what does this version cost?” The main blackjack hand can win while 21+3 loses, or the side bet can win while the blackjack hand loses. The wagers are mathematically separate even though they use some of the same cards.
That separation is the most important thing to understand. Basic blackjack strategy does not make 21+3 a better bet. The value of 21+3 comes from its paytable, number of decks, and exact hand definitions.
The three cards used by 21+3
The side bet normally uses:
- the player’s first card;
- the player’s second card;
- the dealer’s exposed upcard.
No later hit card, double card, split card, or dealer hole card is used for the ordinary 21+3 wager.
Once those first three visible cards are known, the side bet can be settled. The blackjack hand then continues under normal blackjack rules.
That creates results such as:
- 21+3 wins, blackjack later loses;
- 21+3 loses, blackjack wins;
- both win;
- both lose.
Do not combine the two outcomes when deciding whether the side wager has a good price.
What usually counts as a winning 21+3 hand
Modern versions commonly pay for three-card poker categories such as:
- flush — all three cards share a suit;
- straight — three consecutive ranks;
- three of a kind — all three cards share a rank;
- straight flush — consecutive ranks in the same suit;
- suited three of a kind — the same rank and suit across multiple decks, on paytables that distinguish it.
Some older or alternate versions use fewer categories or a flat payout. Some branded variants add progressive or Xtreme features.
The category list printed on the felt or paytable is the contract. A player should not assume that the 21+3 paytable remembered from another casino applies here.
The side bet does not change blackjack strategy
Suppose the player receives 7♠-8♠ and the dealer shows 9♠. The three visible cards create a straight flush for 21+3.
That side bet can be paid immediately according to the table. But the blackjack hand is still hard 15 against dealer 9 and should be played under the applicable blackjack strategy.
The large side-bet win does not turn 15 into a stand. Likewise, losing the side bet does not make the player “due” to win the blackjack hand.
If you need the base-game decision rather than the bonus wager, use Basic Strategy.
A common five-tier paytable
One widely analyzed modern structure pays:
| 21+3 result | Example payout |
|---|---|
| Suited three of a kind | 100 to 1 |
| Straight flush | 40 to 1 |
| Three of a kind | 30 to 1 |
| Straight | 10 to 1 |
| Flush | 5 to 1 |
| Anything else | Loss |
This table is an example, not a universal 21+3 schedule. The house edge changes when any payout changes, when categories are added or removed, or when the number of decks changes.
That is why a side-bet name alone is not enough information to judge it.
Probability and payout have to be evaluated together
A high payout can look attractive because the event is rare. The correct question is whether the payout compensates for the probability.
For an event with probability (p) and a win paying (b) to 1, the contribution to expected return is:
[ p \times b ]
A losing outcome contributes its probability multiplied by -1 unit.
For a paytable with several winning categories:
[ EV = \sum_i p_i b_i - p_{loss} ]
where:
- (p_i) is the probability of winning category (i);
- (b_i) is that category’s payout to 1;
- (p_{loss}) is the probability of losing the wager.
The house edge is:
[ \text{House Edge} = -EV ]
when EV is expressed per unit wagered.
An eight-deck worked example
For one widely used eight-deck 100-40-30-10-5 version, published combinatorial analysis gives approximately these probabilities:
| Result | Probability | Pays to 1 | Return contribution |
|---|---|---|---|
| Suited three of a kind | 0.0244% | 100 | +0.0244 |
| Straight flush | 0.2063% | 40 | +0.0825 |
| Three of a kind | 0.5169% | 30 | +0.1551 |
| Straight | 3.0947% | 10 | +0.3095 |
| Flush | 5.8841% | 5 | +0.2942 |
| All losing results | 90.2736% | -1 | -0.9027 |
Adding the contributions gives an EV of about:
[ -0.0370 ]
or a house edge of roughly 3.70% for that specific version.
The calculation is not transferable to another paytable without recalculating the returns.
Hit frequency is not the same as house edge
Using the same eight-deck example, the total probability of any listed winning hand is about:
[ 1 - 0.902736 = 0.097264 ]
So the side bet wins on roughly 9.73% of resolved wagers, or about one in ten on average.
That does not mean the return is close to 90%. A bet can hit fairly often and still have a meaningful house edge because most outcomes lose one full unit while the winning categories are priced below fair odds.
Hit frequency answers “How often do I get a payout?” House edge answers “What percentage of action is expected to be lost in the long run?” Those are different measurements.
One $5 chip can add a surprising amount of hourly cost
Suppose the side bet has a 3.70% house edge and the player wagers $5 every hand for 60 hands.
Total side-bet action is:
[ 5 \times 60 = 300 ]
Theoretical side-bet loss is:
[ 300 \times 0.037 = 11.10 ]
That is about $11.10 of expected loss from the side bet over those 60 hands.
The $5 chip feels small because it is smaller than the main blackjack wager. Repetition is what turns it into meaningful action.
If the player also has a $25 main wager with a much lower house edge, the side bet can contribute a disproportionate share of expected loss despite using fewer dollars per hand.
That effect is explored more broadly in How Side Bets Change Your Real House Edge.
“To 1” versus “for 1” matters on side-bet signage
A payout of 10 to 1 means the player receives 10 units of profit plus the original wager back.
A payout of 10 for 1 means the total return including the original wager is 10 units, which is equivalent to 9 to 1 profit.
Casinos and game documentation normally specify the convention, but players should read carefully because side-bet EV calculations require the actual profit amount.
This is one reason the Blackjack Payouts page distinguishes settlement language from the visual size of a paytable number.
Deck count changes duplicate-card categories
In one physical deck, it is impossible to have three cards of the exact same rank and suit because only one copy of each specific card exists.
In a multi-deck shoe, suited three of a kind becomes possible because the shoe contains multiple identical copies of cards such as 7♣.
That is why a 21+3 paytable can include “suited trips” only in a game using enough decks to make the event possible. Deck count also changes the exact probabilities of ordinary trips and other categories.
So a paytable copied from an eight-deck game should not be analyzed with one-deck probabilities.
Why the side bet feels better than its expectation
21+3 combines three features that are psychologically powerful:
- the result is known quickly;
- recognizable poker hands create a strong visual pattern;
- rare results can have large displayed payouts.
That makes wins memorable. A straight flush is more vivid than a long sequence of $5 losing chips.
The arithmetic, however, counts every wager equally. If 90% of bets lose in a particular version, those ordinary losses still dominate the action even though they are less memorable.
This is why session tracking should record the side bet separately from the base game. See Blackjack Session Tracking.
Side-bet variance is much higher than the main game’s normal rhythm
The main blackjack hand has many even-money outcomes, pushes, doubles, and occasional 3:2 naturals. A tiered 21+3 side bet concentrates return into rarer events with larger multipliers.
That produces a more uneven result distribution. Two players with the same amount of 21+3 action can finish far apart because one hits a high tier and the other does not.
High variance does not change house edge. It changes the spread of short-term outcomes around the expectation.
The distinction is covered in Blackjack Variance Explained.
A paytable comparison should use EV, not the top jackpot number
Suppose Paytable A offers a 100-to-1 top award while Paytable B offers 50 to 1. It is tempting to conclude that A is automatically better.
But if B pays more on flushes, straights, and three of a kind, its overall expected return could be better despite the smaller headline payout.
The correct comparison is:
- list every winning category;
- calculate or obtain each category probability for the deck count;
- multiply each probability by its profit payout;
- subtract the losing probability;
- compare total EV.
The biggest printed number is marketing information. The whole paytable is mathematical information.
Card counting and 21+3 are separate technical problems
A normal blackjack count is designed around the effect of high and low cards on the base game. That does not automatically tell you whether a poker-style side bet has become favorable.
An advantage player would need a side-bet-specific model that tracks the ranks and suits relevant to 21+3 outcomes. A casual claim such as “the count is high, so play the side bet” is not enough.
For ordinary players, the safe assumption is that 21+3 remains a negative-expectation side wager unless a verified analysis of the exact paytable and current composition says otherwise.
Do not use a side-bet win to change the main-hand strategy
Consider a $25 blackjack hand plus a $5 21+3 wager. The three-card side bet wins $50 on a straight. The player then faces a normal hit-or-stand decision on the blackjack hand.
The $50 win is already determined. It should not make the player more willing to take an incorrect stand, double, or split on the main hand.
Likewise, five losing side bets in a row do not make the sixth wager more likely to win if the dealing process has not created a relevant composition change.
Keep the two ledgers separate: poker-style side bet versus blackjack decision.
How to inspect 21+3 before betting
Before placing the chip, read:
- the exact qualifying hands;
- the payout for each category;
- whether payouts are “to 1” or “for 1”;
- the number of decks if it affects the analysis;
- whether the table is ordinary 21+3, Xtreme, progressive, or another variant;
- the minimum and maximum side-bet amount.
Then compare the wager with the side-bet analysis, not with the main blackjack house edge.
The Blackjack Side Bets Overview gives a broader framework for comparing 21+3 with Perfect Pairs and Lucky Ladies.
When a small side bet does not fit the budget
A player can have a sensible $25 main-game budget and still create unnecessary cost by automatically adding $5 or $10 of 21+3 every hand.
If the side bet is primarily entertainment, decide its budget separately. One option is to limit the number of side wagers rather than making it an automatic attachment to every hand.
Do not increase the side bet after a losing streak to “get it back.” A larger next wager does not repair the EV of the previous wagers.
Category order matters when hands overlap
A three-card straight flush is also, by ordinary description, both a straight and a flush. A paytable does not normally pay all three categories at once. The hand is classified at its highest applicable listed category.
The same logic applies to suited three of a kind in a multi-deck shoe. Three identical 7♣ cards are also three of a kind, but if the paytable has a special suited-trips tier, that higher tier controls.
This matters when checking an EV table. Probabilities should be mutually exclusive: every possible three-card result belongs to one paying category or to the losing group. Adding overlapping probabilities would double-count some hands and produce a false return.
Suited trips explains why 21+3 is a shoe-game side bet
A standard 52-card deck contains only one 7♣, one Q♥, one A♠, and so on. Therefore three identical rank-and-suit cards cannot occur in one deck.
With eight decks, the shoe contains eight copies of each exact card. That makes a result such as 7♣-7♣-7♣ possible when the player’s first two cards and the dealer upcard come from different physical decks inside the shoe.
This is an unusual feature compared with ordinary poker, where one deck is assumed. It also illustrates why casino side bets must be analyzed under their actual equipment configuration rather than by borrowing probabilities from a home-poker intuition.
One paytable change can move the house edge materially
Take the example eight-deck schedule where a flush pays 5 to 1 and occurs with probability about 0.058841.
If a casino changed only that flush payout from 5 to 1 down to 4 to 1, the expected return would drop by approximately one unit of payout multiplied by the flush probability:
[ 0.058841 \times 1 = 0.058841 ]
That is about 5.88 percentage points of return removed from the wager, assuming every other probability and payout stayed the same.
The example shows why a player cannot say “21+3 is around 3.7%” without reading the paytable. A single common-category payout change can move the edge dramatically.
It also explains why a small difference in a frequent lower-tier payout can matter more than a spectacular change to an extremely rare top award.
Combined blackjack and side-bet cost should be calculated as two wagers
Suppose the player bets $25 on blackjack and $5 on 21+3 for 60 hands.
Assume, only for illustration, that the base blackjack game has a 0.50% house edge under the player’s strategy and the side bet has a 3.70% edge.
Base-game action:
[ 25 \times 60 = 1{,}500 ]
Base-game expected loss:
[ 1{,}500 \times 0.005 = 7.50 ]
Side-bet action:
[ 5 \times 60 = 300 ]
Side-bet expected loss:
[ 300 \times 0.037 = 11.10 ]
Combined simplified theoretical loss:
[ 7.50 + 11.10 = 18.60 ]
In this example, the smaller $5 side wager creates more expected loss than the $25 main wager. The reason is the higher percentage edge repeated every hand.
The figures are illustrative and exclude extra base-game action from doubles and splits. The lesson is general: compare each wager’s action and edge separately before adding them together.
A progressive 21+3 wager needs a jackpot-meter calculation
Some installations add a progressive jackpot related to 21+3. A progressive cannot be evaluated from the fixed paytable alone because part of the return depends on the current jackpot amount and the probability of the jackpot-triggering hand.
The generic structure becomes:
[ EV = EV_{fixed} + p_{jackpot} \times J ]
where:
- (EV_{fixed}) is the expected return from fixed awards;
- (p_{jackpot}) is the jackpot-event probability;
- (J) is the jackpot value measured in units of the wager after accounting for the exact rules.
A rising meter can improve the wager’s return. It does not automatically make the bet positive expectation. You need the trigger probability, contribution structure, reset amount, and exact progressive rules.
That is a different calculation from ordinary fixed-pay 21+3 and should not be mixed into a flat-pay house-edge figure.
The dealer upcard creates shared-card dependence, not a second blackjack decision
The dealer’s upcard is simultaneously part of the 21+3 three-card hand and part of the information used for the player’s blackjack strategy. That shared card can make the two outcomes feel connected.
They are connected through the physical cards, but the wagers are still settled by different rules. A dealer 6 may be favorable information for the base blackjack decision while forming no paying 21+3 hand. A dealer 10 may help complete a 21+3 straight while creating a difficult base-game decision.
The correct workflow is to settle or mentally close the side bet, then make the blackjack decision from the blackjack chart. Do not let the excitement of the poker pattern rewrite the base-game strategy.
21+3 is a branded side bet with multiple implementations
Galaxy Gaming describes 21+3 as combining the player’s two blackjack cards with the dealer upcard to make a three-card poker hand, and notes that Xtreme and progressive variations are available. See the official Galaxy Gaming 21+3 product page.
That is useful for defining the game family. It does not tell you that every installation uses the same paytable or house edge. Casinos can offer approved variants with different payout structures.
Published mathematical analyses show how widely the edge can move
The Wizard of Odds 21+3 analysis documents multiple historical versions with different paytables, deck counts, and house edges. The range is wide enough to make one lesson unavoidable: you cannot evaluate 21+3 from the name alone.
A flat-pay version, a five-tier version, an Xtreme table, and a progressive version are different wagers even if the betting circle says 21+3.
The maximum side-bet limit is part of risk control
Casinos often set a side-bet maximum lower than the main blackjack maximum. That protects the game from excessive exposure to rare high-paying outcomes and helps keep chip settlement manageable.
For the player, the same idea is useful on a personal scale. A side bet with high variance and a higher house edge should not automatically grow whenever the main wager grows.
If the main blackjack bet is raised from $25 to $50, there is no mathematical requirement to double the 21+3 wager from $5 to $10. The wagers have separate risk profiles and separate expected costs.
Bottom line
Blackjack 21+3 is a separate three-card side bet using the player’s first two cards and the dealer upcard. Its value is determined by the exact paytable, deck count, and hand definitions—not by blackjack basic strategy. A common eight-deck 100-40-30-10-5 version has been analyzed at about a 3.70% house edge, but other versions can differ substantially. Read the whole paytable, calculate expected value rather than focusing on the top prize, and track the side bet separately from the base blackjack hand.
Operational controls when a table offers 21+3
The side wager uses cards that are already part of the blackjack deal, but it still creates a separate settlement process. Staff need to know which side-bet position belongs to each hand, when side betting closes, the approved hand-ranking order, the exact paytable, whether only the highest category pays, and the verification procedure for unusual premium hands.
The wager also changes total action. A table advertised at a $10 blackjack minimum can carry materially more money per round when several players add $5 or $10 side bets. That extra action should be rated and supervised as a separate wager rather than treated as if it changed the expected value of the main blackjack hand.
A dispute should be resolved from the approved rules and the actual cards and wager state, not from a generic memory of “21+3.” Commercial implementations can differ in paytable and category treatment.