How about a life-insurance policy that can grow a bit when the stock market does? How about getting your capital-gains tax deferred because it's sheltered in an insurance contract?

If that sounds irresistible, you're red meat for a salesperson offering "variable life insurance," the fastest growing form of coverage on the market today.

It's a policy that combines insurance and savings. The savings portion is invested in a number of insurer-offered mutual funds: a stock fund, bond fund, zero-coupon Treasuries, money-market fund, mixed fund, even real estate. You can move your investment from one fund to another, and you pay no current taxes on your interest, dividends or capital gains.

If your investments rise by a sufficient amount, the policy's death benefit rises, too. If not, it will fall -- although many insurers guarantee that your family will always be paid a specified minimum.

But there are angles to this policy that make it a lot less attractive:

Sky-high costs. In the first year of New England Life's Zenith policy, a 40-year-old man might pay 40 percent of his fixed, annual premium in up-front fees and charges (57 percent, if he's age 25). In years two through four, he'd pay 18 percent, and after that, 14 percent. The corresponding costs for Prudential's Variable Life contract: 62 percent in the first year, 16 percent in years two through four and 8 percent after that.

Charter National's no-load (no sales charge) Life inVest policy charges nothing up front, but picks up money in other ways. There's a surrender charge if you quit the policy during the first eight years. Charter charges more for insurance than its competitors. And its investments have to make greater gains before your death benefit will rise.

Shackles on your gains. Cash values rise and fall with the investment return. But sales and administration costs take a sharp cut out of your returns.

Zenith's stock fund was a star performer last year, rising 98 percent. Yet a 40-year-old, buying in that year, would have realized 19 percent on his initial investment, after paying up-front costs -- less, after paying mortality and expense charges. Prudential's fund rose 15 percent; a new buyer who put up $757 would be left with $340 after up-front costs.

A hypothetical 12 percent annual yield over 20 years on Merrill Lynch's Prime Plan (which is higher than the market's historical return) drops to 9.5 percent after paying all the fees.

Molasses-slow rises in death benefits. Few buyers realize that, with many insurers, the policy's death benefit doesn't always rise when the market does. Typically, your investments have to increase by more than 4 percent a year before your death benefit will go up (except for Charter, which ties its policy more closely to changes in the market).

Even then, the rise can be tiny. After five straight years of 8 percent gains from stock investments, the death benefit on a Zenith policy might rise only 0.9 percent. On a Charter policy, it can take more than 25 years of 8 percent gains before the death benefit will rise. What's the big deal in that?

Costly losses. If the gain on your investment is less than 4 percent in many policies, your death benefit can decline. So your insurance can shrink even when the market is inching up. Furthermore, your investments then have to rise by enough to offset the policy's prior underperformance before the death benefit can rise again.

With Charter and policies like it, your policy can lapse if investments go sour and the cash value drops to zero -- unless you put more money in.

Loans from the policy are tax-deferred. But they can permanently reduce the value of your insurance, even if the loan is repaid in full -- so your costs are much greater than they appear. Any borrowing may defeat your hope of letting the stock market push the death benefit up.

It's especially pointless to buy this policy and invest the cash values in money-market funds or bond funds, Glenn Daily, an insurance analyst at Seidman Financial Services in New York, told my associate, Virginia Wilson. That's because they won't yield enough, after costs, to make a significant difference in the value of your policy, and might even pull the values down, he said. He sees only one use for variable life: It suits the person with money to spare, who won't borrow from the policy, who invests the whole cash value in stocks (expecting high returns to put the death benefit up), and then leaves the proceeds to his or her heirs.