BOSTON -- In the 17 years since he hung out his shingle in Cambridge, Graham Gund has forged an unmistakable, idiosyncratic architectural style. At the same time, using his personal fortune as a base, he has built up a formidable, and equally adventurous, career as a real estate developer.

Gund's related activities are of more than passing interest in Washington, because last spring the 47-year-old architect astonished local skeptics by winning a competition to design and develop the Lansburgh's project near Pennsylvania Avenue downtown.

When it's done, two years or so from now, the building Gund has designed -- colorful, picturesque and big -- is likely to be controversial. But it wasn't the architecture that turned heads last year. It was the can-do essence of his idea: Alone among the five entrants, Gund eschewed building any offices at all. Instead, after meeting standards for parking, retail stores and the arts, he proposed that all remaining space in the huge project be devoted to housing.

This decision, so counter to conventional wisdom about the downtown market, produced a plan that greatly exceeded the minimum residential requirement established by the Pennsylvania Avenue Development Corp., which sponsored and ran the competition. More than any niceties of Gund's design, it probably gave his firm a decisive edge. Lansburgh's will be the first residential project to come on line in the area controlled by the PADC, a prospect that without question has given a boost to other, stalled residential projects in the vicinity.

"Everybody else tried to stuff in offices," he recalls, "but it was clear to me that housing is what they {the PADC} really wanted."

Victory was sweet. In 1978 Gund's firm had entered, and lost, the PADC competition to restore and add to the Willard Hotel. "We were just a bunch of kids when we entered those competitions {the Willard and the development of Long Wharf in Boston harbor}," says Peter Madsen, a Gund partner now in charge of the Gunwyn Co., the firm's development arm, "and we ended up with nothing. Maybe we were a little out of our league."

Gund is a big, regular guy with a cleft chin, domical forehead and clear, steady eyes. He dresses like a banker but seems to pay more attention to his ties than your average banker does. He walks through his spacious East Cambridge studios with an impressive, slightly impatient air of command. He more or less scoffs at a widely held notion that 30 or so architects is ideal for a design-oriented firm controlled by the artistic vision of a single architect, admitting only that "the ideal size would be a little smaller" than the 80, mostly young architects currently in his employ. They uniformly snap to when the boss is around, and probably when he isn't.

The firm, he says, is "big enough to do any project and small enough to concentrate" on high-quality design. The work force is divided into teams for specific projects and they'll meet with Gund "once a week or 10 times a day if it's necessary." Before the first meeting with any team Gund will have worked out the basic approach he wants to take, the "concept of how spaces are going to work together" and, presumably, since color and surface pattern have become such major factors in his architectural vocabulary, at least a rudimentary idea of how the building or buildings will look in elevation.

"We're architects first and very much developers second," Gund has said, explaining the relationship between his two businesses. "We don't bother with straightforward development projects. We look for unusual projects -- projects where the architecture matters." Although he oversees the development company, he concentrates on design, leaving most of the deals and their myriad details to Madsen and six assistants in the Gunwyn Co.

Something of the brash spirit of the enterprise is captured in Madsen's statement: "You have to have an idea of how a city will grow so that you can get in and develop properties before they become too expensive. While we do marketing studies, we do them mainly to prove to lenders what we already know will work."

The Gund operation is situated in Bulfinch Square, a restored complex of 19th-century buildings, mostly courthouses (one designed in 1814 by Charles Bulfinch), that in itself is an advertisement for the firm at its ingenious best in architecture, preservation, urban design and development. The very existence of Bulfinch Square has been called "a miracle" by Boston Globe architecture critic Robert Campbell. Aptly so, for it sits next to a blunt, incredibly ugly 1960s high-rise called the Middlesex County Courthouse. The complex was of course slated to be demolished for more of the same.

Gund and company cannot be given sole credit for saving this valuable cluster -- in a turn one somehow would like to think typical of Boston, a political scandal brought a halt to the courthouse expansion -- but they saw an opportunity where almost no one else did. When inducements such as federal preservation tax credits and Urban Development Action Grants became available, they were ready and able to put a complicated development package together.

And they became, after the project was completed in 1984, their own chief tenants. The firm has nearly doubled in size since then and the architects work in splendid, high-ceilinged spaces that used to be courtrooms. The difficulty in convincing more conventional tenants that such unusual spaces are suitable should not be underestimated, but Gund eventually was able to do this, too. Today lawyers are in (apparently) happy habitation of the old court chambers not occupied by architects (or the one utilized by a community arts group, which was part of the package).

It's typical of Gund's projects, too, that this one was a leader in its geographical area -- other restorations are underway nearby, and abandoned industrial plots in East Cambridge, once scorned as useless, are today being transformed into a lively mix of residences, offices and stores.

The architecture is a canny mix of pure restoration, necessary adaptations and new additions -- walls that in certain places replicate existing details and in others are self-confident rephrasings of historical motifs. But the best touch of all was the surgical removal of a few older pieces that had obscured the charm and cohesiveness of the 19th-century cluster. The results of these diverse manipulations are exciting interior spaces and fine buildings confronting a bricked courtyard -- an architecturally framed space that has become a compelling, urbane place.

Combined interests in architectural history, in preservation, in urban connections and urban places, in what makes cities cities, have been constants in Gund's life as an architect. When he was still a student at the Harvard Graduate School of Design he helped to stir up interest in saving Boston's threatened old city hall, and he used his own money to become codeveloper of the project to convert it to commercial offices. Today it's an attractive anachronism, a neat little 1865 piece in the shadow of skyscrapers.

Gund doesn't talk about money, but his supply of it remains plentiful -- his inherited wealth has been estimated at upwards of $100 million. He was one of six children whose father, George Gund, was chairman of the Cleveland Trust Co. and an industrialist. Graham Gund did his undergraduate study at Kenyon College in Ohio and then followed his muse to the Harvard GSD, from which he received master's degrees in architecture and urban design in 1968 and 1969, respectively. (The building that now houses the GSD, designed by Australian architect John Andrews and completed in 1971, was built mainly with Gund family money and is called, appropriately, Gund Hall.)

But it took a while for Gund to make his architecture reflect his interests in history. His design style and philosophy have progressed in phases from the minimal sort of modernism he was taught at the GSD to the exuberant, extremely personal brand of postmodernism one sees today on the drawing boards in his offices -- a metamorphosis that is in many respects characteristic of the talented architects of his generation.

Gund, of course, doesn't like the label "postmodernist" -- understandably, given the vulgar currency of a term applied to every speculative office building with a pediment. But he is, in any larger definition, a postmodern architect. He may owe his penchant for thoroughness in design to Walter Gropius, the Bauhaus expatriate who was his chief mentor at Harvard. But otherwise he has rejected most of the principal tenets of modern architecture -- he likes color (and then some), ornament, pattern, traditional as well as 20th-century materials, defined places rather than neutral spaces, multiple meanings rather than singular ones, urban context instead of the tabula rasa and, perhaps above all, suggestions (but not copies) of history in his buildings.

A key early work was his 1976 renovation of a late 19th-century police station to house Boston's Institute of Contemporary Art, a natural commission in that Gund is the city's preeminent collector of 20th-century art. He left the rough stone Richardsonian exterior intact but gutted the interior to form an open central space, organized in lively fashion by a sequence of diagonal stairwells, and several severe, neutral exhibition spaces.

This straightforward exercise in contrast -- old outside, new inside -- was followed by a more venturesome, but still somewhat tentative, amalgam of new and old in projects such as the Schoolhouse Condominiums, completed in 1980. This was the first of the smaller, "more manageable" design-development efforts Madsen says the firm undertook after those failed ventures in the big time in the late 1970s. The strategy, not then commonplace, was to convert an abandoned schoolhouse in an economically marginal Boston neighborhood into 21 upscale condominiums.

The finely detailed housing units won praise, as did the lobby design, notable for its color and whimsy, with partitions shaped and painted green to suggest trees, chromium railings shaped to suggest vines -- very much in the manner of the "abstract figuration" that Princeton architect Michael Graves was assaying at the time. The conversion won several design awards and, as Madsen is quick to point out, the $2 million project also made "some nice money" for the development firm.

However, besides Bulfinch Square, the most important project of Gund's transitional period was the multiaward-winning Church Street Condominiums, completed in 1983. Again acting as his own developer, Gund here took the burned out relics of a delicate neo-romanesque church in the Back Bay -- the tower and two gabled bays -- and transformed them into frontispieces for a 43-unit apartment building. (The tower itself became one of the more vertical residences anywhere -- seven stories, very few windows.)

This is a terrific essay in contrast, irony and contextual fit. The new building, with its bricks and bays and windows echoing commonplace Back Bay themes, forms an appealing, if angular, cup for the picturesque ruins. But it's not neutral background -- it feels comfortable, looks composed, and yet there's a peculiar, late-20th-century tension in its subtle brick patterning. This collage of familiar motifs made fresh by unexpected juxtaposition probably is Gund's first wholly successful effort to enliven the fac ades of his buildings with color and pattern -- a 19th-century tactic he's pushed to more systematic extremes in his newer buildings, such as Lansburgh's. It'll become the trademark of his mature career.

The Church Street project brought Gund his first avalanche of national attention and, obviously, helped launch the dramatic expansion of his two businesses. More impressive than the number of projects the architecture firm is busy with -- only a small fraction of which are Gund-developed -- is the variety. Only an elegant 31-story office structure nearing completion in Boston comes close to being a skyscraper, which is a standard measure of superstardom in the profession. Nor is this a typical, speculative, single-use development. It combines a setback tower with an extended six-story base that will go far to unite a disparate architectural surround in the booming downtown and, not coincidentally, will do much to enhance pedestrian activity on the streets it faces.

The roster of projects also includes a condominium apartment building on Copley Square opposite McKim, Meade & White's distinguished Boston Public Library; a hotel-office project on a Boston wharf not far from the one they lost out on a decade ago; a total remodeling of a lumpy 1960s office structure at a key downtown intersection; a seven-story retail building next to Faneuil Hall; performing arts centers for a college in North Carolina and for Gund's former boarding school in Connecticut; a natural history museum in Georgia; a grand-style condominium hotel in the New Hampshire hills; a library extension for a Massachusetts village; and an entire center, including residences and stores, for a New England college town.

Most of these buildings will be characterized by Gund's exuberant, eclectic mature style. Although some already hate it -- "His buildings tend to look like toys," comments one young Boston architect -- none would gainsay that Gund's is one of today's more fascinating practices. He has distinguished himself in the complexity of architectural tasks he has undertaken, and the built results so far suggest that he's mastered most of them.

The Lansburgh's project, in any case, is a complicated mix of restoration, alteration and new construction, of residences, commerce and the arts that plays perfectly to Gund's strong suits. There's already much to admire, not least the willingness of Gund and company to take the long view and their simple conviction that, in Madsen's words, "any sort of new town downtown would attract a lot of very different types of people."

There are always questions. One cannot help wondering how the interior spaces will feel when fleshed out in three dimensions -- but the floor plans are demonstrably excellent, providing handsomely for stores, residences and two subterranean theaters (one of 500, another of 200 seats, each emptying into a skylit lobby). Double-loaded residential corridors (with half of the units facing city streets, the other half an interior courtyard) will open onto units that, though tight, exceed Washington norms. Gund, Madsen and colleagues seem confident that a substantial number of two-bedroom and bedroom-and-den units -- approximately half of the 370 total -- will pay off in the long run. The other half will be one-bedroom apartments -- no studios -- with rents starting at about $750 (in today's dollars).

The big esthetic issue is how will it look? Based on the elevations I've seen it'll look fine, if odd -- there's something a bit arbitrary about the exaggeratedly "residential" profiles at the top of the building on Seventh Street. And a good deal of strangeness (such as the inclusion of two transplanted fac ades) was required by the PADC. Pure preservationists won't like it, and many a good architect will say it's too much -- too much decoration, color, fanciful history and so on. Nonetheless, this will be a standout building in a place that desperately needs one now, and in a few decades it'll be at worst an oddity we'll all love and at best a heralded landmark. Maybe both.