Blackjack

Blackjack Insurance

By PayoutCompass · 2 June 2026 · Updated 16 June 2026 · 12 min read
In Short

Insurance is the side bet offered when the dealer’s up-card is an ace, paying 2:1 if the dealer has blackjack. The short answer for almost everyone: never take it. The long answer involves probability math, the “even money” trap that is the same bet in disguise, the one specific exception for card counters, and a handful of myths that get insurance taken thousands of times a day across every casino on earth. This guide is the complete reference.

What Insurance Actually Is

When the dealer’s up-card is an ace, before any other play happens, the dealer pauses and asks: “Insurance?” You can place a separate side wager up to half your original bet. The bet pays 2:1 if the dealer’s hole card turns out to be a ten-value card (giving them a blackjack); it loses if the hole card is anything else.

The mechanics are straightforward. The bet is independent of your main hand. If the dealer has blackjack, your insurance bet wins, your main bet loses (unless you also have blackjack, in which case you push), and the hand is over. If the dealer does not have blackjack, your insurance bet loses immediately, and you play out the main hand normally against the dealer’s revealed cards.

The pitch is intuitive: “insure against losing to a dealer blackjack.” The math says the pitch is wrong for almost everyone. The pitch is also the same pitch under a different name when the dealer offers “even money” on a player blackjack, covered below.

§

The Math, Step by Step

Insurance pays 2:1, which means for the bet to be mathematically fair (zero edge to either side), the dealer would need to have blackjack one time out of every three when showing an ace.

The actual probability in a standard 6-deck shoe at the start of a fresh shoe: there are 312 cards total, 96 of which are ten-value cards (4 tens, 4 jacks, 4 queens, 4 kings, times 6 decks). You have seen three cards before insurance is offered: your two cards plus the dealer’s ace. Assuming none of those three cards were tens (the most common case), the remaining shoe has 96 tens out of 309 cards.

That gives the dealer roughly a 31.0% chance of having blackjack when showing an ace, in the most common starting situation. Compared to the 33.3% break-even threshold, the gap is the house edge:

~31%
Probability of dealer blackjack (6 deck, neutral shoe, no tens in player hand)
33.3%
Break-even threshold (2:1 payout requires this)
~7.4%
House edge on every insurance bet taken

For every 100 ringgit wagered on insurance over time, the casino expects to win about 7 ringgit, 40 sen. Compared to the main blackjack game’s house edge of around 0.5% with basic strategy, insurance is roughly fifteen times worse per dollar wagered. It is one of the worst common bets in the casino, hiding inside a game that has one of the best.

§

Even Money: The Same Bet, Different Costume

When you are dealt a natural blackjack and the dealer shows an ace, the dealer will offer “even money”: a guaranteed 1:1 payout on your blackjack, taken immediately, instead of waiting to see whether the dealer also has blackjack (which would push your hand).

Even money feels different from insurance. It is not different. The math is identical: taking even money is mathematically equivalent to taking maximum insurance on your blackjack hand. The casino has just repackaged the same bad bet with a more reassuring name.

Walk through the math. If you decline even money:

  • Dealer has blackjack (~31% chance): your blackjack pushes the dealer’s, you get your bet back (net: zero).
  • Dealer does not have blackjack (~69% chance): your blackjack pays 3:2, you win 1.5 times your bet.
  • Expected value: 0.69 × 1.5 = 1.035 times your bet.

If you take even money: guaranteed 1.0 times your bet. The expected value of declining (1.035) is higher than the guaranteed even money payout (1.0). Take the gamble; the math is on your side.

The “but you might push” part of the pitch is real (you will push roughly 31% of the time and feel briefly disappointed), but the math says the 69% chance at the full 3:2 payout outweighs the 31% chance at zero. Decline even money. Always, unless you are a counter and the count is at +3 or higher (covered below).

§

When to Take Insurance

For non-counters: never. There is no main-hand situation that makes insurance the correct play. Not when you have hard 20, not when you have a blackjack, not when you are “feeling lucky,” not when the dealer “always seems to have blackjack.” The 7.4% house edge applies to every insurance bet you take regardless of your main hand.

For counters: take insurance when the true count reaches +3 or higher. At true count +3, the shoe contains enough additional ten-value cards that the dealer’s blackjack probability crosses the 33.3% threshold and the bet becomes positive expected value. This is the most valuable single play on Don Schlesinger’s Illustrious 18 deviations list, accounting for roughly 30% of the total deviation value.

The counter precondition matters: you have to actually be counting. Insurance “feels right” because the dealer is showing an ace, but feel has no relationship to whether the count is at +3. The deviation only works as part of a complete counting system, which we cover in our card counting guide.

!

“Insurance for a good hand” is still a bad bet

A common reasoning at the table: “I have a 20, so I should insure it because it would suck to lose this hand to a dealer blackjack.” The 20 is irrelevant. Insurance is a separate bet with its own probability, paying out based only on whether the dealer has blackjack. The strength of your main hand does not change the math on the side bet. Taking insurance with a 20 is the same -7.4% expected value as taking it with hard 16.

§

The Five Myths That Get Insurance Taken

Insurance gets taken constantly across every casino because of a handful of persistent misconceptions:

Myth 1: “It is insurance, like car insurance.” Car insurance has positive expected value because the insurance company spreads risk across millions of policies and prices for tail events. Casino insurance has negative expected value because the casino prices for its own profit. The name is marketing, not analogy.

Myth 2: “The dealer is showing an ace, so blackjack is likely.” Roughly 31% likely in a neutral shoe, which is less than the 33.3% the payout would require to be fair. Likely-ish is not likely enough.

Myth 3: “I should insure my good hands to protect them.” The strength of your main hand is irrelevant to whether the dealer has blackjack. Insurance pays based only on the hole card; your 20 versus your 12 makes no difference.

Myth 4: “Taking even money locks in a sure win.” Mathematically equivalent to taking insurance, which is mathematically negative EV. The “sure win” of 1:1 is smaller than the expected value of letting the hand play out at 3:2.

Myth 5: “Card counters take insurance, so it must be good.” Counters take insurance only when the true count is +3 or higher (perhaps 5 to 10% of the hands where insurance is offered). Counters decline insurance the other 90%+ of the time. The selectivity is the whole game.

§

Insurance With a Player Blackjack

This is the most common situation that prompts even money and insurance offers. You are dealt an ace and a ten-value card, the dealer shows an ace, and the dealer asks: “Even money?” or, equivalently, “Insurance?”

The two are mathematically the same bet on the same outcome. The math:

ActionIf Dealer Has Blackjack (~31%)If Dealer Does Not (~69%)Expected Value
Take even moneyWin 1.0x betWin 1.0x bet1.000x bet
Decline even moneyPush (win 0)Win 1.5x bet1.035x bet

Declining is worth about 3.5% more in expected value. Over enough hands, that 3.5% becomes meaningful money. Take the swing; decline even money on player blackjacks unless you are counting and the count justifies otherwise.

§

Variations on the Standard Rule

Most casinos worldwide offer insurance under standard rules (2:1 payout, available only when dealer shows ace, peek for blackjack before main hand plays). A few variations exist:

  • European no-peek rules: The dealer does not check for blackjack until after the player has finished acting. Insurance is offered before the dealer’s hole card is dealt, and the bet resolves later. Mathematics is essentially the same; the timing differs.
  • Some online operators offer “smart” insurance: A button that takes insurance only when the count would justify it. These are usually marketing features rather than genuine advantage, since the operator’s RNG reshuffles every hand and there is no count to track.
  • Spanish 21 insurance: Works identically but the deck composition (no 10s) changes the math slightly. The Spanish deck has fewer ten-value cards, so the dealer’s blackjack probability with an ace up is lower than in standard blackjack. Insurance is even worse in Spanish 21 than in standard blackjack. Covered in our Spanish 21 guide.
§

Why Casinos Offer Insurance

The honest answer: because it makes money. Insurance bets carry a 7.4% house edge on every dollar wagered, and the bet feels emotionally compelling enough that players take it frequently. From the casino’s perspective, every insurance taken is a small win expected over the long run.

The bet also serves a behavioural function. Players who take insurance tend to think of themselves as making prudent, defensive plays. This framing keeps them comfortable at the table and reduces the emotional impact of dealer blackjacks (which they “insured against”), encouraging longer sessions and more total wagering. The casino benefits twice: directly from the insurance bets and indirectly from extended session length.

Industry observers sometimes describe insurance as the “tax on optimism.” Players who believe they can sense when the dealer has blackjack take it; players who understand the math do not. The casino collects from the first group consistently.

§

Online vs Live Insurance

Mechanics are identical: the option appears, you click yes or no. The math is identical: 7.4% house edge under standard rules.

The practical difference is decision speed. At a live table, the dealer’s offer creates social pressure (the dealer is waiting, other players are watching). Online, the decision is between you and the screen, and the timer (if any) is generous. Online players have more cognitive space to remember “always decline” and act on it; live players occasionally cave to the social moment and take insurance against their own better knowledge.

Both contexts deserve the same answer: decline, unless counting (which is essentially impossible online due to reshuffling at RNG and poor penetration at live dealer).

§

Frequently Asked Questions

Should I take insurance if I have a really good hand?

No. The strength of your main hand is irrelevant to the insurance bet. Insurance pays based only on whether the dealer has blackjack, which depends on the hole card, not on what you are holding. A 20 and a hard 13 have exactly the same insurance EV: negative 7.4% of every ringgit insured.

Why is even money the same as insurance?

Mathematical equivalence. Both bets pay you a guaranteed (or near-guaranteed) amount in exchange for giving up the chance at a larger amount. Working through the probabilities, taking even money on a player blackjack and taking maximum insurance on a player blackjack produce identical expected payouts. The casino renames the bet to make it feel different, but the math does not change. Decline both unless you are counting and the count justifies otherwise.

Doesn’t the dealer always seem to have blackjack when they show an ace?

No, they have blackjack roughly 31% of the time. The 69% of hands where they do not are memorable as “the time I didn’t insure and was fine”; the 31% where they do feel painful and stick in memory. The asymmetry creates the impression of “always,” but the actual frequency is roughly one in three.

If insurance is so bad, why is it allowed in the game?

Because the casino designed it to be there. Insurance is profitable for the house (7.4% edge per dollar wagered) and behaviourally useful (it keeps players engaged and emotionally comfortable). The bet exists for the casino’s benefit, not the player’s. The math is not hidden; it is just rarely advertised.

Are there counting systems that make insurance profitable?

Yes, in the sense that any standard card counting system tracks when the shoe is rich enough in tens for insurance to become positive expected value. Using Hi-Lo counting, the threshold is true count +3 or higher. At that count, take insurance; below it, decline. This is one of the highest-value plays in the entire Illustrious 18 deviations system.

What about “even money” specifically when I have a blackjack?

Decline. The expected value of declining (1.035x your bet) is higher than the guaranteed even money payout (1.0x your bet). The 31% chance of pushing on a dealer blackjack feels worse than it costs; the 69% chance of collecting the full 3:2 payout outweighs it mathematically. Take the swing.

Where to Go from Here

For most players, the insurance rule is the simplest in blackjack: do not take it. For everything else about the side bets layered onto blackjack tables, our side bets guide covers 21+3, Perfect Pairs, Lucky Ladies, and the rest. For the one exception (insurance for counters at high true counts), the card counting guide and the Illustrious 18 deviations walk through the full system. For where to play with rules that minimise the house edge across the whole game (not just insurance), our best online blackjack casinos for Malaysia page lists the operators we rate most highly. Back to the full Blackjack guide.

Written by the PayoutCompass team. We may earn a commission when you sign up through links on this site — it never affects our ratings or what we report. How we make money.

18+ only. Gambling should be fun, not a way to make money. If it stops being fun, get help here.