(MONEY Magazine) – Chances are that if you've mustered the courage to look at this year's ''simplified'' 1040 tax packet, you are probably ready to seek professional help -- either from a tax preparer or a mental-health therapist. The tax reform law has added new complexity to the old forms while creating another whole set of befuddling worksheets for you to complete. So if you have any itemized deductions at all, you should probably see a tax practitioner. But beware: the pros are plenty baffled themselves. A MONEY examination of 50 tax preparers nationwide shows that the new tax law is causing all sorts of problems for those who make their living filling out tax returns. What's more, an easier test we conducted reveals that the IRS' own taxpayer assistors are alarmingly ignorant about the revised tax code (see the box on page 136). In January, MONEY mailed the financial profile of a hypothetical family to the participants in our project, who agreed to prepare the family's tax returns based on the information sent. (They were also encouraged to call us if they needed additional information.) Formulated with the help of S. Theodore Reiner, senior manager with the Washington, D.C. office of Ernst & Whinney, the profile raised complex tax issues relating to the new law, but included no unfair trick questions. Here are the most striking results of our survey: -- The 50 pros computed 50 different amounts of tax due from the family. Their bottom lines varied by as much as 50%. -- The preparers are still uncertain about the nuances of the new tax law. -- In cases where the law itself is unclear, the pros are making up rules of their own that differ greatly. Because there are an unprecedented number of gray areas, it's more likely than ever that you could end up paying a higher tax bill -- or risking an audit -- with some practitioners than with others. -- Pro tax preparers, even the most expensive ones, are not immune from making careless mathematical mistakes. -- H&R Block outlets are capable of handling a difficult return as skillfully as some Big Eight firms -- at a fraction of the cost. While the Block preparers would have charged an average of only $267 to our hypothetical family, the national accounting firm's fees averaged $1,567. Originally, we sent our hypothetical case to 60 practitioners. Fifteen were chosen from each of four categories: certified public accountants with national offices, who charge anywhere from $50 to $250 an hour; C.P.A.s with local or regional firms, who on average charge slightly lower rates; enrolled agents -- independents who have either worked for the IRS as revenue agents for at least five years or passed a two-day IRS test -- who charge between $50 and $150 an hour; and H&R Block preparers, who charge $8 to $24 per tax form -- which works out to $50 for a typical return. Nationwide, there are about 375,000 C.P.A.s, 40,000 H&R Block preparers and 24,000 enrolled agents. Of the 96.1 million personal tax returns filed in the U.S. for 1986, about 44.1 million -- or 46% -- were filled out by professional preparers. Fifty of the 60 practitioners who agreed to take MONEY's test sent back completed 1040 forms to us within two weeks as promised. Seven C.P.A.s with national firms and three H&R Block preparers backed out, in most cases claiming that they didn't have enough time. Because the law is imprecise about several matters pertaining to the hypothetical case, there was no right or wrong bottom-line answer. (No withholding was assumed, for simplicity's sake.) Ernst & Whinney's Reiner, who helped us develop the hypothetical case, anticipated that the typical response would be around $9,000. He was right on the money: the actual average turned out to be $9,105. But the final tax due that the preparers came up with spanned an enormous range, from a low of $7,202 all the way up to $11,881. Of the 50 participants, 16 computed a tax that differed from the $9,105 average by more than $500. (Each participant's fee and bottom line is listed in the table on page 135.) The factor most responsible for the wide variation in the responses was the new, more complicated version of the alternative minimum tax, which will affect more taxpayers than in previous years but still less than 1% of all personal returns. The AMT, as it's called, is a tax computed separately from the 1040 form that prevents taxpayers with sizable deductions from paying very little to the government. Many, but not all, of the preparers who came up with a low tax due simply miscalculated the AMT by omitting an important figure. Five of the preparers who computed an unusually high tax bill did so because they assumed -- quite reasonably -- that a technical corrections bill containing an important AMT change will soon become law retroactive to 1987. Because their fees are relatively high, the five could save their clients as much as $700 by eliminating the need for an amended return after the bill passes. In the meantime, however, the hypothetical family would owe a tax that is $1,596 higher than it would be under the law as it now stands. The lesson: if your preparer tells you that you are subject to the AMT, make sure he spells out the costs and benefits of figuring your return under both the current and the corrected rules. In addition to the AMT, there were three other important reasons for variations in tax calculations. First, some preparers were unfamiliar with the tax code's finer points. A problem in the hypothetical case involving an excess contribution to a 401(k) retirement plan, for example, stumped six participants. In addition, almost half of the practitioners were unfamiliar with new provisions enacted last December concerning mutual fund expenses and master limited partnership income. The errors in such instances involved small dollar amounts, though similar mistakes on your return could conceivably cost you a lot of money. We did not count as errors cases in which participants made assumptions on their own that justified a different answer. These independent assumptions caused some minor variations in the bottom lines. Mathematical mistakes or blatant oversights were the second main reason for differing tax calculations. Thirteen of our participants -- five local C.P.A.s, three enrolled agents, three Block preparers and two C.P.A.s with national firms -- made errors that could be characterized as blunders. The lesson: always review each line of your completed return with your preparer. If he seems vague or uncertain about a particular item, have him show you the pertinent tax code provision or regulation. Remember, if you are ever audited, you -- not your tax preparer -- will be the one subject to back taxes, interest and penalties. The final major cause for differing bottom lines is the gaps in the tax law itself. For some situations, there are no clear rules. In our hypothetical cases, the preparers had to cope with such ambiguities to determine the family's interest and miscellaneous deductions. Aggressive preparers tried to save as much tax for the family as they could justify. But in so doing, they increased the risk that the IRS would audit the family's return. Conservative preparers were more inclined to let the family pay higher taxes. Our participants were split fairly evenly between these two philosophies. C.P.A.s with national or local firms tended to be either very aggressive or very conservative. The enrolled agents and H&R Block preparers were apt to be middle of the road. Not surprisingly, the C.P.A.s, who by and large serve wealthy individuals and corporate clients, charged fees that far exceeded those of the other preparers. The average bill was $1,567 for those at national accounting firms and $966 for the local C.P.A.s. In contrast, enrolled agents and H&R Block preparers, who cater to taxpayers of modest means, charged an average of $582 and $266, respectively. The average tax bill that they calculated for the family was slightly lower than the amount that the C.P.A.s computed: $9,076 for the enrolled agents and $9,027 for the H&R Block preparers, compared with $9,086 for the national C.P.A.s and $9,205 for the local ones. One reason for their high fees is that C.P.A. firms generally offer much more sophisticated year-round tax-planning services. In fact, many C.P.A.s who participated sent along tax-planning tips for the hypothetical family with their tax returns, even though we didn't ask for any. In some cases, their ideas would have saved the family more than the $2,500 that the most expensive C.P.A. charged. No enrolled agent or Block preparer included planning points. The performance of the relatively inexpensive H&R Block preparers was quite good. Considering that the hypothetical family's finances were far more complex than those of the typical Block client, the chain provided excellent value. Many of the Block participants consulted the firm's eight-person troubleshooting team in the firm's home office in Kansas City to ask about particular changes in the law, a step that preparers in branch offices often take if a client has a complicated problem. A group of five enrolled agents came up with bottom lines that were within $9 of one other, an aberration considering how widely the other responses varied. Patricia Burton of Gales Ferry, Conn., one of the five, explains that she and several of the others discussed some of the more challenging aspects of the case. A former president of the 6,000-member National Association of Enrolled Agents, Burton says she frequently consults with other enrolled agents about her client's tax problems. Because there are relatively few enrolled agents, the group tends to be more closely knit than other professional preparers, she says. Some of the most interesting results of our test pertain to the specifics of our hypothetical case. The profile we sent describes the Johnsons, a couple earning a combined salary of $100,000 with three children, each of whom earned enough in 1987 to be subject to tax. The family's investments included stocks, corporate and municipal bonds, mutual funds, limited partnerships and U.S. Savings Bonds and Treasury bills. The Johnsons moved during the year and kept a second home, which they rented out part of the time. To complicate things further, they also took out a second mortgage on the second home. After reviewing the returns, Ernst & Whinney's Reiner said that aside from the AMT, the area that posed the biggest problem was the mortgage interest deduction on the second home. Says he: ''Calculating the AMT and the interest deductions was brutal.'' Here, from the top of the tax return to the bottom, is a summary of how the preparers computed the Johnson family's taxes:

SALARIES, INTEREST AND DIVIDENDS Aside from a couple of mathematical errors, the practitioners generally had little trouble with this section. One exception: six preparers failed to report as taxable interest $125 that was earned on an excess 401(k) contribution the wife made. Several others also didn't count the money but made assumptions on their own that justified the omission.

CAPITAL GAINS On this line, 20 practitioners incorrectly counted $400 in mutual fund fees as taxable income. They would have been correct before December, but since then Congress has passed a bill saying that such so-called phantom income should not be counted as income in 1987.

LOSSES FROM LIMITED PARTNERSHIPS Another 21 of our participants, some of whom knew about the mutual fund fee change, missed a different provision in the December tax bill. They were not aware that tax-shelter losses cannot be used to offset fully income from a master limited partnership, a type of partnership that is publicly traded like a stock -- a fact that some of our practitioners didn't know.

STATE AND LOCAL TAXES Here, only a couple of preparers made an error. Nevertheless, the responses differed. Forty-four of our participants aggressively followed a 1983 appellate court ruling to determine the deduction while six adhered to more conservative IRS rules. The issue concerned the amount of property-tax expense on the family's second home that could be deducted on Schedule A, their itemized deduction form. The IRS and the court disagree about the proper way of allocating expenses between personal and rental use when a second home is rented out part of the time. Using the court ruling formula added $150 to the Johnsons' personal property tax write-off compared with the IRS approach.

INTEREST DEDUCTIONS For this write-off we received 49 different answers. The preparers essentially had no guidelines because the IRS hasn't yet written final regulations explaining how to deduct interest on a second mortgage on a second home that is leased out part of the year. Because the second mortgage was used for personal expenses and exceeded the original cost of the property, not all the mortgage interest was deductible. The most aggressive of our participants ended up writing off a total $21,022 in interest, while the most conservative deducted only $17,320. Most of the responses ranged between $18,100 and $18,900.

MISCELLANEOUS DEDUCTIONS Here, 38 preparers deducted a conservative amount while 12 were more aggressive. The controversy related to a four-year subscription to a trade publication that cost $200. Following IRS regulations, the conservative practitioners deducted only one-fourth of the $200. The others wrote off the full amount, explaining that there's no law actually prohibiting the $200 write-off. In an audit, the extra $150 would most likely be disallowed, which would leave it up to the Johnsons and their preparer to decide whether to take their case to tax court. Considering the small amount, they would probably just pay up. The results of our project are anything but reassuring. They provide a graphic warning to all taxpayers during this filing season: don't sign your return until you have carefully scrutinized every line. That's true whether you filled out the forms yourself or hired the most expensive accountant in town. But don't hesitate to hire someone to do the job for you. Despite the problems they had with our hypothetical case, the majority of the pros proved that they can save taxpayers a lot of anguish, and maybe even a fair amount of money.

BOX: Testing the IRS

SO HOW COME THE IRS KNOWS ALL THE ANSWERS WHEN YOU GET AUDITED?

When the General Accounting Office examined the IRS telephone assistance program a year ago, it found that the agency answered only 79% of the GAO's questions correctly. The IRS vowed to do better this year and has launched a major public relations campaign promising to ''put the service back in The Service.'' It's a nice thought. But a MONEY survey conducted this past January -- using easier questions than those we asked of the tax practitioners -- suggests that the IRS ''helpers'' have gotten worse. In fact, the assistors were right only slightly more than half the time. Forty percent of those we queried, for example, didn't even know that the IRS itself, with great public fanfare, had waived the penalty on tax underpayments for workers who had too little withheld from their paychecks last year. Maybe the IRS should call H&R Block. MONEY placed a total of 100 calls to IRS assistors. Each helper was asked one of 10 questions, prepared with the help of Ernst & Whinney, and each question was posed to 10 different assistors. The results were appalling. Overall, only 55 responses were correct and 45 were inaccurate. In four cases, the assistors said that additional research was needed and that a written response would be sent within eight working days. Did research help? No way. All four written answers were wrong. Spokesperson Johnelle Hunter defends the IRS by saying that MONEY's questions were ''not typical.'' Larry Endy, the director of the GAO's study, says the IRS made the same claim last year. Says he: ''We asked questions that we thought the IRS people should be able to answer whether they received them 10 times or a million times.'' So did MONEY. The assistors were most knowledgeable about the new kiddie tax (MONEY, February 1988) and the reduction in basic deductions for such expenses as union dues. All the representatives we asked knew that a 13-year-old child with $100 in bank interest and $700 in paper route income must file a 1987 tax return. And nine of 10 correctly responded that miscellaneous deductions, including union dues, can be written off only to the extent that their total exceeds 2% of adjusted gross income. But 60% were not aware that the cost of sending a child to an overnight summer camp can be counted toward the child-care credit in 1987. Also, seven of 10 didn't know that interest paid on a loan used to buy tax-exempt bonds is not deductible. In the wake of last year's GAO report, the IRS hired 1,000 more assistors, bringing the total to 4,500. The service also expanded its training sessions for telephone representatives and required its supervisors to monitor more phone calls. In addition, to solve last year's problem of constant busy signals, the number of phone lines has been expanded by 30%. Unfortunately, despite those efforts, IRS helpers remain woefully ill-informed about the new tax law. G.A. Jr.

CHART: TEXT NOT AVAILABLE CREDIT: NO CREDIT CAPTION: More than ever, the tax preparer makes all the difference The 50 tax practitioners -- of four types -- who completed returns for our hypothetical family came up with different tax totals -- and fees. The average tax (black line) was $9,105, but the returns ranged from $7,202 to $11,881. The average fee was $779, but it varied from $187 to $2,500. DESCRIPTION: See above.