In the 1980s, eye-popping returns on many mutual funds made sales charges seem a small price to pay for a share of the action. In the more sober 1990s, however, sales charges and expenses will likely eat up a larger portion of profits. While performance should always be the focal point for potential investors, a thorough examination of the fee schedule in a fund's prospectus can have more of a payoff than in the past.

On the surface, funds with lower sales charges, or loads, and lower expenses seem to have an obvious edge. But the issue isn't so cut and dried. Funds with high loads can outperform no-load funds, and funds with high expenses can produce better returns than funds with low expenses. So while keeping an eye on loads and expenses is important, investors shouldn't be obsessed with buying the cheapest fund they can find.

COMPLEXITY. Figuring out what impact the sales charges and expenses have on your return is getting more difficult. Along with traditional front-end-load and no-load varieties, more funds have added what some in the industry call a spread load--officially it's a"12b-1 fee"--to the mix of expenses. It shaves a set percentage off the fund's total return each year for a set number of years. The fee is often coupled with a deferred sales charge, or back-end load, which may be up to 6% in the first year but gets lower the longer a fund is held. You pay a back-end load only if you cash out of the fund.

Fee schedules are tougher to decipher today, but there is some good news on the load scene: Of the 2,400 funds tracked by Chicago's Morningstar Inc., only 30 or so carry the maximum allowable load of 8.5%. The bad news is that, while pure no-load funds still exist, many old-line no-load funds are becoming low-load funds, with sales charges of 2% to 3%. And expenses have been rising, too.

Choosing what type of sales charge you prefer to pay for a fund--or whether you're willing to pay one at all--depends first on whether you have the desire and time to do the work of picking a mutual fund on your own. If you use a financial planner or a broker, you're likely to pay between 4% and 6% of the total invested assets for their advice. Some say that can be money well spent: "You're paying for an extra level of service, for someone to help you answer tough questions, such as 'What is your real risk tolerance?' " says Don Phillips, publisher of the guide Morningstar Mutual Funds. If you decide to research mutual funds on your own, you might as well buy no-loads, he says.

An investor's time horizon is also important in figuring out whether paying a sales charge makes sense. Of course, if you pay a large load and stay in a fund only a year, the load will represent a bigger part of your return than it would if it were spread out over a greater number of years. "If it's a period shorter than five years, I'd pay attention to the sales charge," says A. Michael Lipper of Lipper Analytical Services. "If it's beyond five years, I would look at it but not be terribly influenced."

A look at the load-adjusted performance of front-end-load and no-load funds shows no-load funds with a definite edge in the performance derby. In the aggressive growth category, the top-performing fund for the year ending July 31 was a no-load, the tiny $5 million American Heritage fund--even with its mammoth 6.7% expense ratio. The top three funds, before adding in sales charges, include a no-load, a fund with a 1% back-end load, and a fund with a 5.5% front-end load. Adjusted for load, however, the top three include two no-loads and a back-end-load fund.

The preponderance of no-load funds among the top performers is even more noticeable in bond funds. The top seven funds in the government-bond category, based on one-year total returns, are all no-load funds. Before factoring in a load, three load funds make it into the top ten. But when you adjust for the loads, they all take a big dive in the rankings.

Take the TNE Government Securities fund, for example. Its 18.78% total return vaulted it into the top ten performing government-bond funds based on one-year returns. Factor in the 4.5% load, however, and it slips to No. 52, with a 13.43% total return.

If you rely mn mutual funds for income and you find a bond fund you like that has a load, choose a front-end load over a 12b-1 fee, recommends Phillips. You'll have a fatter income stream because front-end funds have lower expenses, and expenses come out of that income stream.

DIRECT HIT. Many people who are willing to pay a load--but who shy away from paying 5% or so up front--find the pay-as-you-go option appealing. But the fees have critics: "My problem with 12b-1 fees is that they can put pressure on the manager to boost yield and increase the risk," says Phillips. Unlike front-end sales charges, which come directly out of invested assets, 12b-1 fees take a big bite out of yield, since they come out of total return.

In the past, fund companies have charged as much as 1.25% in annual 12b-1 fees. That could make a 12b-1 fund an expensive proposition for a longtime investor. Recently, however, the Securities & Exchange Commission set a 0.75% cap on 12b-1 fees, effective next July. A new 0.25% service fee brings the actual cap up to 1%. Total sales charges are capped at 6.25% for a fund that charges the service fee and at 7.25% for those that don't. For long-term investors, that makes the choice between a front-end load or the pay-as-you-go option more or less of a moot point, says Phillips. Some funds have created new classes of shares to switch their existing shareholders into once the 12b-1 fee roughly equals the front-end load offered on the "sister" fund.A hypothetical example shows how pay-as-you-go plans start off with an edge over front-end-load funds but lose that edge over time. Consider two funds, each with similar objectives, portfolios, and performance. One fund has a 5.75% front-end load. The other has a 12b-1 fee of 0.75% and a back-end load that starts at 3% and declines by 1% every two years.

Say you invested $2,000 in the front-end fund. The fund's 5.75% load would take a $115 bite, leaving you with $1,885 working for you. An investor in the 12b-1 fund puts the entire $2,000 to work.

Assuming a 10% return, the front-end-load investor would have $2,073.50 after one year in the fund. Investors in the 12b-1 plan would have seen their $2,000 grow to $2,183.68 after year one. But what if the investor in the 12b-1 fund had to cash out after a year? The 3% back-end load is levied on the original $2,000 investment, leaving $2,124--still more than the investor in the front-end load.

If the funds continue to earn 10% a year, the hypothetical 12b-1 fund outperforms the front-end-load fund for about eight years. After that, the front-end-load fund takes the lead, with $4,889.20 in assets after 10 years. The 12b-1 investor has $4,815.20.

For those with large sums to invest, front-end loads may have an edge over 12b-1 plans with back-end loads. Front-end loads ratchet down on bigger investments, whereas back-end loads make no such accommodations.

A fund's fee schedule shouldn't be your first stop. Spend some time figuring out the type of fund you want and then examine its portfolio and performance. If, after that, you don't like the fee choices on a particular fund, well, there are still thousands of other funds to choose from.

Before it's here, it's on the Bloomberg Terminal. LEARN MORE