Jami Rubin wasn’t buying it. The influential Goldman Sachs (GS) analyst, sitting just a few rows back from the podium, waited patiently through the corporate cheerleading and then raised her hand for a question: “Can you guys just remind me what the strategic benefits are of keeping the consumer business under the parent organization?” Then she added, “Sorry, I just cannot help myself.”

Enter the Conglomerate Killer. Rubin, who had covered the pharmaceutical industry since the 1990s, had been pushing lumbering corporations to slim down for years. Her campaigns against Pfizer, Abbott, and Baxter had each been met with staunch internal resistance, but the companies had ultimately succumbed, and investors had profited rather handsomely from, as Rubin put it, “the unlocked value.” (According to a Goldman research note in March, Abbott freed $92 billion for shareholders by spinning off its pharma division; Baxter, a more modest $4 billion since doing the same.)

For more on Global 500, watch this video: [fortune-brightcove videoid=5042304089001]

In 2011, in an attention-getting research report titled “Breaking Up Is Easy to Do,” Rubin called upon virtually every major drug company to spin off assets. A year later, when Gorsky took the helm of J&J, the most cyclopean conglomerate in the clodhopping pharma realm, the New Brunswick company shot up to the top of Rubin’s hit list.

Her long campaign got a boost in January when Artisan Partners, a San Francisco–based investment firm, made the case for splitting up the health care company, citing among other sins the poor performance of two of its three divisions—the consumer products and medical device businesses, which account for more than half of J&J’s revenues. (Drugs account for the rest.) “There is absolutely no evidence that these businesses create more value as a result of being together,” says Daniel O’Keefe, a managing director at Artisan, and the lead portfolio manager of its Global Value Fund, which at the time held a tiny (0.2%) ownership stake in J&J. “In fact, the evidence is overwhelming that they perform worse than they would as independent companies,” he says. (Goldman’s Rubin put a number to that thesis in mid-March, estimating that the sum of J&J’s parts was worth $40 billion more than the whole.)

Make no mistake about it, however: J&J isn’t splitting up anytime soon. Gorsky has thought long and hard about his company’s sprawling footprint—and the CEO insists he’s not clinging to it for the sake of legacy, but rather for strategy. J&J’s seemingly unwieldy size is Gorsky’s formula for achieving something rare among publicly traded companies these days: a safe haven that grows.

The first component in that description—stability—has been J&J’s calling card for decades. In April the company raised its dividend for the 54th straight year, and it remains one of only two U.S. companies with a AAA credit rating from S&P. (The other is Microsoft.)

Though J&J’s 2015 revenues fell 5.7% ($4.3 billion) year over year, due largely to unfavorable currency exchange rates (see “Beware the Almighty Dollar”), sales have been inching up since Gorsky became boss. On the profit side the news has been brighter, with a three-year annualized growth rate of 12.4%. On July 19, J&J reported solid sales and profits, given the recent global turbulence, and raised its guidance for 2016 sales by $300 million. During Gorsky’s tenure, J&J’s stock (including dividends) has returned 120% to shareholders, compared with 69% for the S&P 500.

“They’re in a great position,” says Kristen Stewart, an analyst at Deutsche Bank. “They have the luxury of time and the ability to look at different opportunities across different business units. That is what a diversified business platform affords them. I have always considered J&J the master of its destiny.”

If J&J is the master of its destiny, then Gorsky seems well-suited to be the master of it. Square-shouldered and trim, Gorsky looks as if he could have been bred in J&J’s Human Performance Institute. (Yes, one of its many businesses is to train corporate athletes.)

The CEO exercises for 60 to 90 minutes per day, usually at 4:30 a.m., and often in the private gym that abuts his office. He has a photographic memory, “ripped arms” (per his media chief), and the classic good looks of a sitcom dad. He has a clutch of favorite sayings—“I’m pleased, but not satisfied”; “History should be a source of pride, but it can’t be an anchor”—that, delivered in measured tone and manner, have a programmed quality. His employees confirm he is human—a guy who likes to take lunch in the cafeteria and sit with colleagues he doesn’t yet know.

A middle child raised in the Midwest, he enrolled at West Point, where he trained as an engineer, and then served six years in the military, a few of them as an Army lieutenant in Europe, before trying his hand as a drug salesman for J&J in California. He rose quickly through the company ranks, then left in 2004 to become COO at Novartis Pharmaceuticals, then four years later returned to J&J—where, after an amicable bake-off with another contender, he got the top job in February 2012.

His appointment came at one of the darkest periods in J&J’s long history. Five days before he was named the company’s seventh-ever CEO, it recalled 574,000 bottles of grape-flavored Tylenol because of a bottle design flaw. The event would hardly have constituted news had it not been for J&J’s roughly two-dozen recalls in the previous 30 months—a cascading reputational and quality-control crisis that had engulfed the company’s consumer business, a stable of well-regarded brands like Benadryl, Aveeno, and Johnson’s Baby Shampoo.

Gorsky, a fitness buff and triathlete, takes a spin class at J&J.Photograph by Timothy Fadek—Redux Pictures

The issues had been the subject of a government hearing in 2010, but J&J couldn’t stop stepping in it. New egregious lapses emerged almost every month: Fragments of wood and metal were found in Rolaids; excessive bacteria in baby lotion. In several cases, products were recalled because of a foul odor, a problem traced back to shipping pallets. The company also drew scorn after it was found to have executed a “phantom recall” of Motrin—contracting a company to buy all the substandard product off shelves instead of issuing a public notice. In 2011 the FDA placed three of J&J’s manufacturing facilities under consent decree—­essentially government oversight. (J&J acknowledges the failures.)

When Gorsky was named CEO, many believed he would clean up the mess but otherwise bring little change. Gorsky was a consummate insider—a man who, in his two decades at J&J, had helped run the two biggest of its three divisions. But from the start, Gorsky challenged fundamental aspects of the company he had inherited—including the once-sacrosanct principle of decentralization: the notion that the conglomerate’s 250-odd units had to remain wholly separate. For much of its existence J&J had grown largely through acquisition—buying up well-managed companies and letting them be. J&J’s executive team didn’t bother with integration or standardization.

But in Gorsky’s view, decentralized didn’t have to mean disconnected. The whole point of a broad-based health care enterprise was that the different business units could work together to find synergies, cross-fertilize ideas, and reap cost savings that could be reinvested in the business. That, thought Gorsky, was the best way to unlock value.

What he was determined to keep unchanged, hokey as it might sound, was the central role of the company’s credo. Outside J&J’s walls the idea sounds improbably quaint—and the sort of thing corporate HR departments trot out on “New Employee Day” and never mention again. But at J&J the credo was as sacred as a constitution.

Founded in New Brunswick in 1886 by three brothers named Johnson, the company grew slowly for a generation. Then Robert Wood Johnson II decided, reluctantly, to take the family business public. He fretted about the effects that market pressures would have on the company’s values, and as a safeguard he wrote a 307-word statement of corporate principles, spelling out to whom J&J’s responsibility lies: First are patients and physicians; second, employees; communities, third. Shareholders? They come last.

Seventy-three years later that credo—J&Jers typically pronounce it, like the British do, as “cray-dough”—is routinely invoked around the company. It’s inscribed in stone at the headquarters’ front entrance; it hangs in every meeting space; it’s there, supersized, on the wall across from Gorsky’s desk. Until a decade ago, newly hired employees were mailed two copies of the credo—one to take to work, the other to frame at home.

Today there are “credo challenges”—something like a corporate crisis drill, in which a business decision is analyzed through a close reading of the credo—and biennial credo surveys, in which employees evaluate the business according to how well it upholds the standards of the credo. When Gorsky first presented his long-term vision for J&J to the board, it was the credo, he says, that was at its core.

Size, of course, was part of that vision. Given the enormous shifts in the health care industry and the large number of government and institutional customers and partners involved, huge scale could be a rare asset for negotiating deals. Heft and breadth were competitive advantages—as long as one could connect the parts.

It was Sandi Peterson’s job to connect them.

Gorsky knew the Bayer executive from the time they had served together on an industry medical-device board, and once CEO, he wasted little time luring her to J&J. In late 2012, Peterson became group worldwide chairman, a newly created position that held sweeping responsibility for and oversight of the struggling consumer group and technology across the enterprise, among other things.

Left to right: Gorsky, J&J group worldwide chairman Sandra (“Sandi”) Peterson, and CFO Dominic Caruso.Photograph by Timothy Fadek

More unusual than her job title: She was an outsider, an exec who hadn’t climbed the ladder at J&J or been scooped up in an acquisition. Such had been the career paths for much of the company’s senior ranks. But the very nature of Peterson’s mission required an outsider, it seemed.

When Peterson joined, each of J&J’s 250 operating companies had, for example, its own HR policy, its own financing system, and its own procurement process. General managers oversaw all these ­decisions—from who mowed the lawn to which accounting software was used. When employees took jobs with different operating companies, there was no consistency. It was as if they were joining a completely different company. IT across the enterprise was particularly fragmented: Everyone did it his own way. Gorsky had made it a priority to bring order to this unwieldy machine—“sometimes a customer doesn’t want to deal with 250 J&J’s,” he says.

A polished McKinsey alum who also serves on the Microsoft board (MSFT), Peterson has a short blond bob and a pair of cycling pedals installed under her desk. She hits the gym almost every day, and you get the sense, within seconds of meeting her, that she gets shit done. She had led similar efficiency-finding exercises at Bayer and, before that, at Whirlpool (WHR) and Nabisco, and she speaks as someone accustomed and impervious to blowback. “I know how to do this,” she says.

Peterson is feverishly aligning the 250 operating companies on everything from the timing of financial forecasts to employee car policies. Key to the standardization process—which is now more than halfway complete and en route, she says, to saving the company $1 billion—is moving almost all the company’s data to the cloud (in a private, HIPAA-compliant way, she says). By 2018, 85% of J&J’s data will be cloud-based.

“They’ve been very progressive, innovative, and forward-leaning in how they’ve made the move,” says Andy Jassy, CEO of Amazon Web Services (AMZN), of J&J’s effort. Whereas many companies take a conservative, wait-and-see approach on such technology moves, Jassy says J&J is out front in defining what health care compliance in the cloud looks like—and has even gotten Amazon to move faster: “They really pushed us to build a product that will appeal to heavily regulated industries,” he says.

Peterson, who has also brokered partnerships with Google (GOOG), Apple (APPL), and IBM (IBM) on a host of tech projects, rattles off statistics in conversation as if they were “ums” and “ers”: J&J launches 450 apps per year (from the 7 Minute Workout to a bedtime app for babies); it has migrated all of its 120,000–plus employees to a single HR database on Workday; it processes as much data per day as eBay; it warehouses 500 terabytes of data—or “2.5 times as much data as resides in the IRS data warehouse,” she told analysts.

Much of J&J’s recent culture shift started with Paul Stoffels—before Gorsky became CEO. The business that Stoffels inherited in 2009 when he was made J&J’s global head of pharmaceuticals R&D was broken. Patent expirations put billions of dollars of the division’s revenues at risk. “We didn’t have a choice. We could either go fully off the cliff, or we could go partially off the cliff and fight our way back up,” he says.

Many analysts thought the business was not worth salvaging. The pharma division was just as decentralized as any other. Accordingly, J&J’s self-described approach to drug development was “many bets.” Put another way, it was willy-nilly. J&J’s seven different drug R&D organizations operated in completely siloed fashion. In some cases multiple companies pursued the same drug targets, and each had its own system for handling clinical or regulatory development. There was no sense of a common mission: Employees identified proudly with their operating companies—they were a Centocor person or a Janssen person, not a Johnson & Johnson person.

“We didn’t have a choice,” says Dr. Paul Stoffels, J&J’s chief scientific officer. “We could either go fully off the cliff, or we could go partially off the cliff and fight our way back up.”

Stoffels’s first move was to merge the units under his purview into one: Janssen Research and Development. The new umbrella group was organized into five therapeutic areas that would target 11 disease states. The restructuring was painful: Nearly 7% of J&J’s global workforce had their positions eliminated. Twelve of the division’s 25 facilities were shuttered, and nearly 200 projects were slashed, but it provided the business focus and a single operating model.

This new entity had a streamlined development process, a highly coordinated system that Stoffels, a musician, calls Accelerando. Under this model the discovery effort never stops: Global teams—statisticians in China, data managers in India, regulatory folks in Europe—work 24/7 to speed drugs to market. The assembly-line approach has cut months, and in some cases years, off the development time. Stoffels squeezed another two months from the process by zeroing in on how J&J submitted applications to the FDA—“That’s time we controlled; it’s just about using re­sources better”—and sticking a team in a hotel together for one month to find ways to do it better. This is how Darzalex, J&J’s breakthrough multiple myeloma drug, went from Phase I to clinic in just 39 months—years less than the average oncology medicine, according to the Tufts Center for the Study of Drug Development—and went to market just one day after approval.

Since 2009, J&J’s struggling pharmaceutical division has become the company’s star. With $31 billion in revenues in 2015—up 40% since bottoming out in 2009—the division is now handily J&J’s largest. Janssen has also outperformed the industry, topping the IDEA Pharma Productive Innovation Index for four years running. Its 17 drug approvals in 10 years have put it in a league of its own, says Bernard Munos, a pharmaceutical innovation consultant. “No other company has come close to that,” he says. The industry average is 0.8 per year.

Stoffels, though, has pushed far more than the timetable. He has redefined the process for drug discovery at the ­company—joining R&D with S&D (search and development) with I&D (invitation and development). The company doesn’t care where new drug candidates come from—whether discovered in-house, from its own Janssen researchers, or from entrepreneurs and scientists outside the company.

Dr. Paul Stoffels, J&J’s chief scientific officer, photographed at J&J Beerse, Belgium. Stoffels heads up R&D and sets the “innovation agenda” across the enterprise.Photograph by Justin Jin

He set up four innovation centers in biotech clusters—Cambridge, Mass.; Menlo Park, Calif.; London; and Shanghai—around the world, places where scientific entrepreneurs can interact with the company’s own drug and technology scouts. The centers host regular events, featuring experts or startup gurus, and engage with anyone who pings, calls, or drops in. They make available J&J resources—a regulatory expert or an introduction to a possible funding partner, for instance—and give feedback.

“We want a Nordstrom-like level of customer service,” says Jeff Calcagno, a former McKinsey consultant and tech entrepreneur who now works with New Ventures at J&J’s Innovation Center in Menlo Park. “Entrepreneurs aren’t used to that,” says Calcagno. “They’re used to big companies blowing them off.”

J&J’s white-glove touch also extends to dealmaking. From his own experience as a biotech founder (his company developed three major AIDS drugs), Stoffels, who is now J&J’s chief scientific officer, knew drug companies tend to have a rigid, prescribed approach to deals. He thought they could be done more creatively, in a range of forms that better suit the particular partner—maybe J&J would fund an experiment without claiming ownership to the data, for instance. Such flexibility lets J&J work with companies earlier and more casually—a strategy that mitigates risk and builds relationships. Between 2013 and 2015, J&J reviewed more than 3,400 opportunities through these centers. In some 200 of those cases they did deals.

Yet more radical is what the company is doing at six biotech incubators, called JLABS, in the U.S. and Canada. There, company reps do everything possible to help life-science startups succeed, whether by introducing them to a J&J scientist or opening up their compound library. But in this “business model,” J&J gets nothing in return (other than a nominal rent payment). Startups can come, do their science, and leave. They can invite in financiers, including J&J’s competitors, to take a look. There are no stings attached.

The seeming benevolence had initially puzzled some observers and was dismissed by others, but in Stoffels’s vision, the gain for the company is all in the relationship—and embedding in a community of innovators. That, after all, is where good ideas come from.

The JLABS are now incubating more than 140 companies—and there’s a waiting list for new tenants at each locale. One Cinderella tale out of JLABS belongs to Arcturus, a three-year-old biotech founded by two big pharma defectors who quit their jobs in 2013 with $50,000 and a dream to develop RNA-based therapeutics to treat rare diseases. Two and a half years later their company—founded with no assets, technology, or drug pipeline—has scored nearly $2 billion in deals (one worth up to $240 billion to develop hepatitis B medicines for J&J; the other with another company). “That’s tech time in the life sciences,” says Melinda Richter, head of JLABS. “That’s what we’re going for.”

Joseph Payne, an Arcturus co-founder, has no doubt that JLABS is good for scientific progress; he’s less sure it’s a winning strategy for J&J in the long term. “Do they help companies raise money, execute and bring science to an exciting level, do deals with JLABS companies? Yes. But at the end of the day,” he wonders, “will they make money and commercialize all this innovation—or will their competitors swoop in and say, thank you for validating their technology? That’s the mega-question.”

Sure, it’s a radical strategy with an uncertain business model. But what else would you expect from a sprawling, old-fashioned conglomerate?

Check out the full Global 500 at fortune.com/global500 for company profiles, financial data, stock quotes, graphics, and more.

A version of this article appears in the August 1, 2016 issue of Fortune.