A year and a half ago, Sylvan Learning Systems Inc., the thriving Baltimore education company, decided to spin off Caliber Learning Network Inc., a start-up corporate training venture launched in partnership with MCI Communications Corp.

Caliber went public at $14 a share, climbed as high as $18.75 and then went into a long and inexorable slide that has left the stock languishing at less than $2 a share.

Now Sylvan is about to spin off another subsidiary, Prometric Inc. Judging from the poor performance of Sylvan's last spinoff effort, some investors are wondering if they are headed for another one-way trip on the down escalator.

"I think this one is substantially better," said analyst Scott Soffen, who follows the education industry for Legg Mason Inc.

Prometric provides computerized testing services -- qualification tests for architects, cosmetologists, computer programmers, stockbrokers and other professionals, plus all the drivers' license tests in Britain.

"Prometric is a whole different story," Soffen said. "It's a fast-growing, profitable business and it represents about half the company."

Caliber also hit the market looking good, until the company issued its first financial report. Though Caliber insisted the growth of its quarterly loss to $6.5 million from $2.9 million was "in line with expectations," the stock was in single digits within a week and has yet to recover. Even putting its courses on the Internet hasn't helped.

The spinoff of Prometric will not take place until next year and it'll be some time after that before the market decides whether the second chapter of Sylvan's tale of two spinoffs is the best of deals or the worst of deals.

Whatever the verdict, the Sylvan spinoffs demonstrate that the performance of a company's stock can depend as much on executives' skills at financial engineering as on the fundamental success of the business.

The dilemma for stockholders is that Sylvan is a solid, profitable and growing company, but that's not apparent from looking at its stock.

"The educational services industry has been an absolute disaster from a stock price perspective," Legg Mason's Soffen said. "It's very out of favor with Wall Street."

Sylvan's stock peaked at $36.87 1/2 on July 22, 1998. It closed Friday at $16 1/4, down 56 percent from that high.

That's actually not bad, compared with what has happened to the other stocks in what was once one of the Washington region's hottest niche industries:

Computer Learning Centers Inc. stock has lost about 90 percent of its value since February 1998, when it peaked at $37.12 1/2. It closed Friday at $3.56 1/4. Fairfax-based Computer Learning Centers became embroiled in a series of disciplinary actions by regulators and disputes with disgruntled students who complained that they were not getting the quality of computer training they were promised. The damage to the company's reputation has slashed enrollments by almost 25 percent, producing a loss of $3 million in the six months ended July 31.

Vcampus Corp. shares, which soared to $25.62 1/2 on Sept. 19, 1997, when the company was called University Online Publishing, have fallen about 90 percent, closing Friday at $2.31 1/4. Reston-based Vcampus is offering five times as many courses as a year ago, but the company is burdened with some old software and consulting operations that turn off investors. The stock stays down even though the company has cut its first-half loss to $3.5 million from $11.2 million.

Strayer Education Inc. stock has dropped 60 percent to $15 on Friday from $38.50 a share on Nov. 20, 1998. The profit-making accredited college, which awards degrees in business subjects, has grown to 13 campuses from nine, with 11,500 students enrolled this semester. But analysts were predicting enrollment of 12,000 and the expansion cost more than expected, crimping profits. Instead of a per-share profit of $1.41 the company had been expected to earn this year, it will earn closer to $1.21 a share, Salomon Smith Barney Inc. analyst Brent Garcia said in an Oct. 6 update for investors. That means Strayer could become a $20-a-share stock, but not the $31-a-share stock that Garcia had projected.

Even analysts who are confident about the long-term prospects of the education industry admit that the stocks are not likely to rebound -- even if the companies get their acts together.

Though Sylvan's profits have been right on target so far this year, Garcia has reduced his target price for the stock. Aiming for $48 a share last spring, Salomon Smith Barney now says $31 is a reasonable expectation. Garcia maintained that target and a "buy" rating on the stock even after the Prometric spinoff plan was announced.

Salomon Smith Barney is more than a disinterested observer. It's one of several affiliates of Citigroup Inc. that together are Sylvan's largest stockholder, with 18.5 percent of the shares.

The spinoff transaction will take two steps: first, an initial public offering of about 20 percent of Prometric's stock, then distribution of the remainder of Prometric's shares to Sylvan stockholders. Neither the number of shares to be sold nor the price has been determined.

Implementation of the spinoff depends on obtaining a ruling from the Internal Revenue Service that Sylvan stockholders will not have to pay taxes on the Prometric stock they receive. The waiting line for IRS decisions on such cases is long, however, and it could be until the middle of next year before a ruling is obtained.

Because it is so far off -- and could be unraveled by a negative IRS ruling -- the spinoff has not had an effect on Sylvan's stock.

Like Caliber, Prometric is an effort by Sylvan to diversify away from the business for which it is best known: tutoring students. In fact, the company is more like an education, training and testing conglomerate serving clients who range from schoolchildren to graduate students to working professionals.

Those efforts to broaden its business have produced a paradoxical backlash on Wall Street. Investment analysts -- the professional stock pickers who at the top of their game can make $1 million a year -- complain that it is too hard to understand a company involved in such a range of businesses.

"One of investors' main concerns with the company has been its complexity," Salomon Smith Barney's Garcia said in a recent report. "By splitting off Prometric, Sylvan takes a large step toward simplifying its investment story and giving its investors greater visibility into its business."

Separating Sylvan's teaching business from Prometric's testing also would eliminate a conflict of interest and could thereby open doors to new business for Sylvan's teaching operations and Prometric's testing business. Tutoring students and prospective professionals for tests is a growing business, but clients who hire a company to run their testing program do not want the company tutoring students on how to beat those tests. As separate companies, Sylvan and Prometric can pursue testers and test-takers.

Even ignoring the growth potential and the benefits of splitting into companies that even a Wall Street analyst can understand, shares of Prometric should be worth more as a separate stock than they are as part of Sylvan.

That's the premise behind the project: Sylvan's stock price will go down, but theoretically not by as much as the new Prometric stock will be worth.

Prometric is likely to end up being the more profitable of the pair. Its side of the business is expected to produce 60 percent of combined testing and tutoring revenues this year. And as every school kid knows, it's better to be giving tests than trying to pass them.