(FORTUNE Magazine) – There's that one high-school photo--you know the one--where you're the living, breathing definition of cool. You're dressed in your favorite funky psychedelic polyester shirt, bell-bottom pants that go out to there, and a haircut inspired by John Travolta, circa the Welcome Back, Kotter years. Where's that picture now? Likely buried in a box in the attic after that day when your kids got hold of it and passed it back and forth, snickering, "You wore that?"

Cool today, uncool tomorrow. That's what happens to something so entirely subjective. So when FORTUNE decided to revisit past Cool Companies on the fifth anniversary of its list, we knew that while some of the 125 companies we've selected would make us proud, others would look like the nightmare we called our prom date. More than a few did leave us scratching our heads, wondering just what were we thinking. (Massively parallel supercomputers kinda sounded cool, didn't they?)

But we also found that we've been pretty good at identifying key infotech trends early on, like the growth of the Internet, the rise of corporate intranets, and the birth of a little industry called enterprise resource planning (ERP) that has become the best software business not dominated by Bill Gates. That canniness stems from the way we pick candidates. While Cool Companies is by far the least scientific of FORTUNE's lists--any tech company, large or small, public or private, profitable or not, is eligible--it has its own unique rigor. To identify candidates, we talk to scores of the smartest industry people we know, technologists and graybeards and high-tech investors, and find out which companies have them most excited.

While we didn't exactly have the Midas touch--we were looking for businesses whose leaders or technologies had buzz, not great stocks--our picks did fine. If you had invested $1,000 in each of the 42 companies that either were publicly traded when they appeared on our list or had IPOs after we anointed them with Cool Company status, you'd now have $76,830. That's a tad better than what you'd have if you'd invested your $1,000 installments in the S&P 500 ($75,223), and a whole lot better than the returns you'd have enjoyed investing in two indexes that track smaller or more tech-oriented companies: the Nasdaq composite ($69,453) and the Russell 2000 ($60,498; for our methodology, see chart). A handful of our picks have made people downright rich: companies like Peoplesoft (a Cool Company in 1995, with a 74.9% average annual return), Cisco Systems (1994; 86%), and Excite (1995, when it was still Architext Software; 87.6%).

That said, staking money on the companies on this list is investing at its most extreme. While the cachet of being deemed cool has definitely caught on (more than 150 hopefuls sent information this year, and they can all quit calling now), making the list is no assurance of success. More than a few people around Silicon Valley consider this FORTUNE list the kiss of death. Four of our companies went belly up after we featured them; others teeter perilously close to oblivion. A dozen or more have changed strategy so completely that they bear little resemblance to the cool company we fell in love with not so long ago. The problem many of our cool companies face: At some point, even Internet shareholders demand something more tangible than a "coolness factor"; at some point, they'd like your bottom line to be something other than very icy-cold.

Are there lessons investors can draw from the example of FORTUNE's 125 companies? There are. For starters, no matter how cool their products, tech companies that don't stay on the cutting edge are doomed. Cool companies that waited too long to reckon with the rise of the Internet flamed out. "When was the last time you bought a CD-ROM?" asks a chagrined Steve Podradchik, whose Medio Multimedia, a 1994 listee, went broke two years later, a victim of the shakeout in multimedia software. What's Podradchik doing now? He runs Marketwave, a software firm that analyzes Web traffic for corporations.

Then there's Silicon Graphics. When we tabbed SGI back in 1993, there seemed no cooler company on earth. SGI's workstations produced graphics that dazzled Hollywood, providing more gee-whiz for the buck than anyone else's. But that company, too, couldn't adjust to the cutting edge--in this case, cheaper microprocessors and the inroads made by Microsoft's Windows NT.

SGI's no Medio, but the point's the same: When it comes to the kind of tech companies we highlight, timing is everything--for both investor and company. The speed at which ideas and trends spring up and disappear has given birth to a new measure in American culture: Internet time. We listed Netscape as a Cool Company in 1994, weeks after it was founded as Mosaic Communications. At one point its stock was the quintessential example of the riches to be made on the Net. Now that Netscape's on its fourth or fifth strategic shift in three years, it's actually trading below its IPO price. Live by the Net, die by the Net.

Another lesson to be learned is that cool products are no guarantee of success. We've picked a bunch of companies that had wonderful ideas and products but no customers: ideaMarket (1997), for instance, was going to sell all kinds of textual content on its Austin, Texas, Website. It had even figured out a way to efficiently collect the small payments required to buy the posted articles. This seemed a logical use of the Web, of the way the Internet was supposed to disintermediate old-media companies and get information directly to the people. There was just one small problem: No one showed up. Now there's a message on the site: "Sales of these products have recently stagnated, to a point where the ideaMarket website is unable to trade profitably." Virtual I-O (1996), in Seattle, made an incredibly cool pair of 3-D glasses that looked as if they'd be great for game players. But when Virtual I-O tried to sell the $799 gadgets over the Net, it couldn't find consumers willing to pony up the big bucks. A new owner is now trying to market a version of the glasses to dentists, cosmetic surgeons, and small businesses.

Truth be told, the cool companies that did best as stock picks were, well, not so cool. Beneath the flashy surface of the past few years of Internet euphoria, many companies have thrived by making applications that put software and the Net to work for businesses. For the most part, their stocks did well--perhaps because since they were serving businesses, they actually had paying customers. Check out the mundaneness of some of our best stock picks: Security Dynamics (1994, up 98% a year), which makes digital keys that help ensure safe E-commerce, and CheckFree (1996, up 40.5%), which makes software for big financial institutions that want to offer customers electronic banking.

Perhaps the most stultifyingly complex of all the business-oriented applications we highlighted was enterprise resource planning. ERP applications are the software equivalent of mainframes--they seem to take over entire companies. The suppliers that have thrived in this business have thrilled shareholders: PeopleSoft; SAP (1995) is up 65.4% a year; even Baan (1996), which trails the other two, is up 86.4% a year. Will those eye-popping returns continue? Hard to say. Many analysts feel that ERP may have run its course. Remember, we're talking Internet time here.

You'd think that with the benefit of this kind of review, our Cool Companies from now on will all be Internet-related firms selling only to businesses, right? Heck, no. We'll continue to choose companies that the smartest high-tech sources we know say are cool, whether they're selling plumbing supplies over the Internet or developing technology to sort socks after they come out of the dryer. We're not even ready to count out all of our past cool companies that seem down and out right now. Look around you, and consider that boot-cut jeans, polyester shirts, and, of course, John Travolta are all hot. Once something's been "cool," it can always stage a comeback.