KEY TAKEAWAYS
CONTENTS
Insurance in blackjack is a side bet you can make when the dealer’s face-up card is an ace. It costs up to half your original bet and pays 2 to 1 if the dealer’s hidden card gives the dealer blackjack. If the dealer does not have blackjack, the insurance bet loses and the hand continues. Despite its name, insurance does not protect your hand. It is a separate wager on one question: does the dealer have a 10-value card underneath?
For a player who is not counting cards, the answer to whether it pays is no. The bet carries an expected loss of about 7.7% of the amount insured, which is far higher than the house edge on the main blackjack hand. Most basic strategy guides therefore say to decline insurance every time.
This article explains how the bet works, where the 7.7% comes from, how “even money” relates to it, and what a hand looks like with the arithmetic shown. Rules vary slightly by casino, so check the table before you play.
What Insurance Means at a Blackjack Table
Insurance is an optional side bet that the dealer has blackjack. It is offered only when the dealer’s upcard is an ace, because an ace plus any 10-value card (10, J, Q or K) makes a two-card 21. A dealer showing a 10 can also have blackjack, but no insurance is offered in that case at standard tables.
When the dealer shows an ace, the dealer asks, “Insurance?” before play continues. You place the bet on the insurance line printed on the layout, in front of your main wager. The maximum is half of your original bet, so a $10 hand allows up to $5 of insurance. You can also choose to insure for less than half, though most players either insure the full amount or decline.
The bet is resolved immediately, before you act on your hand. If you want a full walkthrough of the rest of the game, see how blackjack is played. In many US casinos the dealer checks the hole card right away, using a peek device or by looking at it, so the insurance result is known within seconds. Some tables, especially outside the US, use no-hole-card rules, and the details differ.
Insurance is a bet on the dealer’s hidden card, not on your hand. That is why a player holding 20 and a player holding 12 face exactly the same insurance decision: the hidden card is essentially equally likely to be a 10-value card either way.
How Insurance Works and Where the House Edge Comes From
The mechanics take four steps. First, the dealer shows an ace. Second, you decide whether to place an insurance bet of up to half your original wager. Third, the dealer checks for blackjack. Fourth, the insurance bet is settled: it pays 2 to 1 if the dealer has blackjack, and it is lost if the dealer does not.
The payout
A 2 to 1 payout means each dollar insured returns two dollars in profit. A $5 insurance bet that wins pays $10, and you keep your $5 stake. If the dealer has blackjack, your original bet normally loses, so the insurance win offsets that loss exactly on a full-size insurance bet.
The math behind the cost
A bet paying 2 to 1 breaks even only if it wins 1 ÷ (2 + 1), or 33.33% of the time. The dealer has a 10-value card underneath about 30.77% of the time using the infinite-deck approximation, because 4 of every 13 ranks (10, J, Q, K) count as 10. Since 30.77% is lower than 33.33%, the bet loses money on average.
The expected value follows directly. Per $1 insured, the bet wins $2 with probability 4/13 and loses $1 with probability 9/13: (2 × 4/13) − (1 × 9/13) = −1/13, or about −7.7%. This is the house edge on insurance for a player who is not counting cards. An insurance bet costs roughly fifteen times more per dollar than a typical 0.5% blackjack hand with good strategy.
Real shoes differ slightly from the approximation, because the cards already on the table change the odds a little. The exact figure depends on the number of decks and what has been dealt, which is why this article states the infinite-deck value.
Worked Example: A $10 Hand and a $5 Insurance Bet
Suppose you bet $10 and are dealt 10♠ and 8♦, a total of 18. The dealer’s upcard is A♥. The dealer offers insurance, and you place $5 on the insurance line.
Case 1: the dealer has blackjack. The hidden card is K♣, so the dealer shows A♥ and K♣. Your $10 main bet loses. Your $5 insurance wins 2 to 1, which is $10 in profit. Net result: −$10 + $10 = $0.
Case 2: the dealer does not have blackjack. The hidden card is 6♦, giving the dealer a soft 17. Your $5 insurance loses, and the hand plays on. If you then win the main hand, you collect $10, but you are $5 lower than you would have been without insurance, so the net is +$5 instead of +$10.
Now compare the long-run cost. Over many hands where you insure $5, the expected result of the insurance bet alone is (4/13 × $10) − (9/13 × $5) = $3.08 − $3.46 = −$0.38, or −7.7% of the $5 insured. You pay about 38 cents on average each time you insure a $10 hand. That is an average over many repetitions, not what happens on any single hand.
Even Money: Insurance by Another Name
Even money appears when you hold a blackjack and the dealer shows an ace. The dealer offers to pay your blackjack at 1 to 1 immediately, instead of the usual 3 to 2, and without waiting to check the hole card. Taking even money is mathematically identical to buying insurance on a blackjack.
Check it with a $10 bet. With a full $5 insurance bet, a dealer blackjack gives a push on your main bet plus $10 from insurance, a $10 profit. If the dealer lacks blackjack, you win $15 on the main bet and lose $5 on insurance, again a $10 profit. Either way you profit $10, exactly what even money pays.
Declining even money gives an average profit of 9/13 × $15 + 4/13 × $0 = $10.38 in the infinite-deck approximation. Even money therefore gives up about $0.38 per $10 bet, or 3.85% of the original wager, in exchange for removing risk on that one hand. It is a guaranteed result, but guaranteed does not mean free.
Common Insurance Mistakes and Myths
The first myth is that insurance protects a good hand. It does not touch your hand at all. The insurance bet and the main bet are settled independently, so a strong 20 gains nothing from insurance, and a weak 12 gains nothing either.
The second myth is that you should insure when you “feel” the dealer has a 10. Feelings and streaks do not change the odds, because the hidden card was fixed before the dealer asked about insurance. Only the actual mix of cards left in the shoe changes the odds, and tracking that is what card counters attempt; a streak or a feeling does not reveal it.
The third mistake is treating even money as a free win. It is a win of $10 instead of an average of about $10.38, so the decision costs money over time. Finally, some players think card counters always take insurance. Counters insure only when the remaining cards are rich in 10-value cards, and counting is not a way to make a living; casinos may ask counters to leave or limit their bets.
Payouts, House Edge and Practical Tips at the Table
The table summarizes the bet. Figures use the infinite-deck approximation, and real shoes vary by a small amount.
| Item | Value |
|---|---|
| Insurance payout | 2 to 1 |
| Maximum insurance bet | Half the original bet |
| Break-even chance needed | 33.33% |
| Chance dealer has a 10-value hole card | about 30.77% (4 in 13) |
| Expected result per dollar insured | about −7.7% |
The table shows the gap between the probability you need and the probability you get. Declining insurance is the standard basic strategy choice, and it saves about 7.7 cents of every dollar you would have insured.
Before you play, check the table placard and the layout for the blackjack payout, since a 6 to 5 table raises the house edge by about 1.4 percentage points and matters more than any one insurance decision. Set your budget before you sit down, as covered in how to buy in at a casino table. If a dealer asks about insurance, a polite “no insurance” is enough.
What to Learn Next
After insurance, the next topics are splitting, doubling and basic strategy, each of which cuts the house edge on the main hand. No betting pattern changes the cost of a bet, a point shown with real numbers in why the Martingale betting strategy fails.
