Blackjack Insurance: Is It Ever Worth Taking?
Insurance pays 2 to 1, and it loses about 7 percent of every dollar you put on it in a shoe game. Here is the exact math, why even money is the same bet in disguise, and the one situation where a card counter takes it.
Published 2026-10-26 · Updated 2026-11-03
How insurance works
When the dealer’s up card is an ace, the dealer offers insurance before checking for blackjack. It is a side bet of up to half your original wager that the dealer’s hole card is a ten-value card (10, J, Q or K).
- If the dealer has blackjack, insurance pays 2 to 1. You lose your main bet (unless you also have blackjack), so a full half-bet insurance breaks you even on the hand.
- If the dealer does not have blackjack, you lose the insurance bet and the hand plays on.
That “breaks you even” is what sells it. It feels like protection. It is actually a separate bet on one question: is the next card a ten?
The math: why it loses
A 2 to 1 payout is fair only if the hole card is a ten exactly one time in three. Win one bet of 2 units, lose two bets of 1 unit: zero. So the break-even ten density is 33.3 percent.
A full deck has 16 ten-value cards out of 52, or 30.8 percent. With the dealer’s ace removed it is 16 out of 51 in a single deck and 96 out of 311 in six decks: still well under one in three.
Insurance EV = 3 × (chance of a ten) − 1. In a fresh 6-deck shoe that is 3 × 0.3087 − 1 = −7.4 percent of the insurance bet.
For comparison, the whole game of blackjack played with perfect basic strategy costs about 0.5 to 0.6 percent. Insurance is more than ten times worse than the game it is attached to.
Exact odds by deck and hand
Your own two cards change the odds slightly, because any tens you hold are no longer in the shoe. Chance the dealer’s hole card is a ten, and the expected value of insurance per dollar wagered, with the dealer showing an ace and no other cards seen:
| Decks | Your hand | Chance of a ten | Insurance EV |
|---|---|---|---|
| 1 | Two non-tens (9,7) | 32.7% | −2.0% |
| 1 | One ten (10,6) | 30.6% | −8.2% |
| 1 | Two tens (10,10) | 28.6% | −14.3% |
| 2 | Two non-tens | 31.7% | −5.0% |
| 2 | Two tens | 29.7% | −10.9% |
| 6 | Two non-tens | 31.1% | −6.8% |
| 6 | One ten | 30.7% | −7.8% |
| 6 | Two tens | 30.4% | −8.7% |
| 8 | Two non-tens | 31.0% | −7.0% |
| 8 | Two tens | 30.5% | −8.5% |
Two lessons. First, every line is negative: off the top of any shoe, insurance never pays. Second, insuring a good hand like 20 is the worst version, because the two tens you hold make a dealer ten less likely.
Insurance calculator
Pick your game, your hand and the Hi-Lo true count. The calculator removes the dealer’s ace and your two cards from the shoe, then shifts the ten density by the count.
The count adjustment is an estimate: each true count point adds about 0.4 ten-value cards per 52 left in the shoe. Off the top it is exact.
Even money is insurance
When you have blackjack and the dealer shows an ace, you are offered “even money”: take 1 to 1 now instead of risking a push. It sounds safe. It is exactly the same bet as insuring your blackjack for half your wager:
- Insure and the dealer has blackjack: you push the hand and win 2 to 1 on insurance. Net +1.
- Insure and the dealer does not: you win 3 to 2 on the hand and lose the insurance. Net +1.
Either way you get +1, the same as even money. Decline instead and, in a 6-deck game, your blackjack is worth 1.039 units on average: about 3.9 percent more than even money. In single deck it is 4.1 percent. Take the risk and decline, unless the count says otherwise.
When counters take it: true count +3
Insurance is the one blackjack bet that depends only on the ten density of the shoe, which is exactly what a Hi-Lo count tracks. When enough small cards have come out, the remaining shoe passes one ten in three and insurance turns into a winning bet.
For Hi-Lo in shoe games that happens at a true count of +3 or higher. It is number one on the deviations chart, and in Don Schlesinger’s ranking in his book Blackjack Attack: the single most valuable index play in blackjack, worth more than any other index play, because you usually have your biggest bet out at high counts and the dealer shows an ace about one hand in thirteen.
| True count | Insurance? |
|---|---|
| Below +3 | No, and decline even money |
| +3 or higher | Yes, take insurance, and take even money |
The rule is the same whether you hold 20, a stiff 16 or blackjack. Insurance is not about your hand. It is a bet on the shoe.
Practice turning insurance down in a free blackjack game.
The Deviations drill deals ace-up hands on both sides of +3 and grades whether you insure, so the index becomes automatic.
FAQ
Should you ever take insurance in blackjack?
Not without counting. Off the top of any shoe insurance loses about 2 to 9 percent of the amount bet, depending on decks and your hand. A Hi-Lo card counter takes it at a true count of +3 or higher, where the remaining shoe is more than one third tens.
Is even money a good bet?
No, for the same reason. Even money is identical to insuring your blackjack. Declining is worth about 3.9 percent more in a 6-deck game. Take even money only at a true count of +3 or higher.
What does insurance pay?
2 to 1. You can bet up to half your original wager. If the dealer has blackjack you win the insurance and lose the main bet, so you break even on the hand.
Why does insurance lose money?
It pays 2 to 1, which is fair only if the dealer's hole card is a ten one time in three (33.3 percent). In a fresh shoe the real chance is about 30.8 percent, so the house keeps the difference.
Does insurance protect a good hand like 20?
No. It is a separate bet on the hole card, and holding two tens makes a dealer ten less likely. Insuring 20 is the worst insurance bet at the table.
More guides: all blog posts · how to count cards · basic strategy chart · deviations chart