The decision you face with a paired hand
You sit down at a blackjack table, buy in, and your first hand comes back as two identical cards. The dealer asks if you want to split. Most players hesitate here, because splitting costs an extra wager equal to your original bet and turns one hand into two. The question is not whether you can split, but whether the rules of the specific table make splitting the worse option.
Pair splitting is one of the few decision points in blackjack where the correct move changes sharply with the dealer's upcard. A pair of eights is almost always split. A pair of tens almost never is. Between those extremes sits a set of pairs where the answer depends on what the dealer is showing and which rule set the table uses.
This article walks through the pairs you should generally leave alone, why the dealer's upcard matters more than your own hand, and where the table's posted rules override any general chart you may have memorised.
What splitting does to the hand
When you split, the dealer separates the two identical cards into two independent hands and places a second wager next to the first. You then play each hand in turn, drawing cards until you stand or bust. The two hands settle separately against the dealer's final total.
Splitting changes the structure of the round in three ways. It doubles the amount of money at risk on that round. It removes the option to treat the pair as a single starting total, such as playing two nines as 18. And it creates two hands that must each be played according to the same table rules, including any restrictions on re-splitting, doubling after split, or hitting split aces.
The arithmetic of the game does not change when you split. The house edge on each resulting hand is the same as on any other hand played under that rule set. What changes is the expected loss on the round, because you now have two wagers in play instead of one. Splitting a pair that should be split reduces expected loss relative to playing the pair as a single hand. Splitting a pair that should not be split increases it.
Pairs that are almost never split
Some pairs are strong enough as a single hand that breaking them apart is a structural mistake. The clearest case is a pair of tens or any two ten-value cards, such as a ten and a face card. A starting total of 20 is one of the strongest positions in blackjack. Splitting it creates two hands that each start from ten, a much weaker position, and it doubles the wager to do so. The same logic applies to a pair of face cards.
A pair of fives is another hand that should generally be played as a single total. Two fives make a starting total of 10, which is a strong drawing hand. Splitting fives creates two hands starting from five, which are weak. Most published basic-strategy charts treat a pair of fives as a ten and never split it.
A pair of fours sits in a similar category. Two fours make an eight, a weak starting total, but splitting them creates two hands starting from four, which are weaker still. The better play in most rule sets is to hit the eight and try to improve it, rather than pay an extra wager to create two poor hands.
| Pair | Why splitting is usually wrong | What most charts advise instead |
|---|---|---|
| Tens or face cards | 20 is already a very strong total; splitting weakens it and doubles the wager | Stand |
| Fives | 10 is a strong drawing hand; two hands starting from five are weak | Treat as 10 and double or hit according to the dealer's upcard |
| Fours | 8 is weak but two hands starting from four are weaker and cost an extra wager | Hit |
Pairs where the dealer's upcard decides
For most other pairs, the correct move depends on what the dealer is showing. The underlying principle is that you split when the dealer is in a weak position and you can afford to put more money on the table, and you avoid splitting when the dealer is strong or when the pair already forms a playable total.
A pair of nines is the clearest example. Two nines make 18, a solid total. Against a dealer upcard of 7, 10, or ace, most charts say to stand, because 18 is strong enough to compete and splitting risks turning a good hand into two mediocre ones. Against a dealer upcard of 2 through 6, where the dealer is more likely to bust, splitting nines can be the better play because it lets you put more money down in a favourable situation. Against an 8 or 9, the decision shifts again and depends on the exact rule set.
A pair of sixes follows a different pattern. Two sixes make 12, a hand that busts easily if you hit. Against a dealer upcard of 2 through 6, splitting is often the preferred move because it gives you two chances to build a stronger hand while the dealer is weak. Against a dealer upcard of 7 or higher, splitting sixes is generally avoided, because the dealer is stronger and you are paying an extra wager to create two hands that still start from a weak total.
A pair of twos, threes, or sevens follows similar logic. The split is more attractive against a weak dealer upcard and less attractive against a strong one. The exact threshold varies by rule set, which is why a single memorised chart can mislead you if the table rules differ from the chart's assumptions.
Why the rule set changes the answer
Published basic-strategy charts are derived from a specific set of table rules. The number of decks in the shoe, whether the dealer stands or hits on soft 17, whether doubling after splitting is allowed, and whether re-splitting is permitted all change the arithmetic of the game. A chart built for a six-deck shoe with the dealer standing on soft 17 and doubling after split allowed will not necessarily match a table with different rules.
This matters most for the borderline pairs, where the difference between splitting and not splitting is small. A pair of nines against a dealer ace, for example, can flip between stand and split depending on whether the dealer hits soft 17. A pair of fours can shift between hit and split in rule sets that allow doubling after split and restrict re-splitting.
The practical consequence is that you should treat any general splitting advice as a starting point, not a fixed rule. Before you play a paired hand, check the table's posted rules and the game information panel. If the operator offers a strategy chart for the specific game variant, that chart is more reliable than a generic one, because it is built from the rule set in use.
Procedural mistakes new players make
The first mistake is splitting tens because the hand looks like two strong cards. It feels aggressive and powerful, but it converts a near-certain win into two uncertain hands and doubles the wager to do it. The expected loss on the round goes up, not down.
The second mistake is treating splitting as a way to recover a losing position. Splitting does not change the house edge on the game. It changes the amount wagered and the number of hands in play. If you split a weak pair against a strong dealer upcard because you are behind and want to win it back in one round, you have increased the amount at risk in exactly the situation where the dealer is favoured.
The third mistake is not knowing the table's split rules before you act. Some tables allow re-splitting, some do not. Some allow doubling after a split, some do not. Some restrict split aces to a single card per hand. The correct decision for a borderline pair can change based on these rules, and you cannot make that decision correctly if you do not know them. The rules are posted on the table or in the game information panel, and reading them before you sit down is part of playing correctly.
What changes and what stays fixed
Splitting decisions change with the dealer's upcard and the table's rule set. What stays fixed is the structure of the game. Blackjack is a comparing card game where the objective is a hand total closer to 21 than the dealer's without exceeding it. The house edge is built into the rules and does not disappear when you split. A correct split reduces expected loss relative to the alternative; it does not create a winning position out of a losing one.
The way to approach any paired hand is to identify what the pair makes as a single total, what the dealer is showing, and what the table rules allow after a split. If the pair already makes a strong total, splitting is usually wrong. If the pair makes a weak total and the dealer is weak, splitting is often right. If the pair is borderline, the rule set decides.
FAQ
Should I ever split a pair of tens?
Almost never. A starting total of 20 is one of the strongest hands in blackjack, and splitting it creates two weaker hands while doubling your wager. The only exception would be a highly unusual rule set or side-bet structure, and even then the general advice is to stand.
Does the dealer's upcard matter when deciding whether to split?
Yes, and it is often the deciding factor. Splitting is generally more attractive against a weak dealer upcard such as 2 through 6, where the dealer is more likely to bust, and less attractive against a strong upcard such as 7 through ace. The exact threshold varies by pair and by rule set.
Where do I find the correct splitting rules for the table I am playing?
The table's posted rules and the game information panel carry the specific rule set in use, including deck count, dealer stand rules, re-splitting, and doubling after split. If the operator publishes a strategy chart for that specific game variant, use that chart rather than a generic one, because it is built from the rules in play.
Play responsibly
Gambling should be treated as paid entertainment, never as a way to earn income or recover losses. 1-800-MY-RESET (1-800-697-3738)
18+ or 21+ depending on where you are; follow the minimum age that applies to you.
This article is general information about how these mechanics work. It is not legal advice and not a recommendation to gamble or to use any particular operator. Availability and legality differ by jurisdiction — check the rules that apply where you are.