How to Identify Antique Furniture: Dating a Piece by Its Construction

Two chests of drawers sat ten feet apart at a Bucks County estate sale, both tagged $600, both labeled "Chippendale" in felt-tip. Serpentine fronts, bracket feet, ball-and-claw hardware: cousins, at a glance. I pulled the top drawer from each and flipped it. The first showed four uneven hand-cut dovetails with a knife-scored baseline. The second showed identical scalloped pins in rounded sockets, a machine joint that did not exist before 1867. One chest could be late-1700s. The other was post-Civil War, dressed up to look older.

That is the whole method. To identify antique furniture, you read the construction, not the silhouette: dovetails, saw marks, nails, screws, secondary woods, and maker marks each carry a datable signature, and the true age is set by the newest original evidence on the piece. Style tells you what a piece is pretending to be. Joinery, tool marks, and hardware tell you when it was actually built. When a monumental Chippendale desk-and-bookcase from the Nicholas Brown family, made in Rhode Island around 1760 to 1770 and attributed to John Goddard, sold for $12.1 million at Christie's in 1989, it was construction and documented history that separated it from ten thousand Victorian copies of the same form.

Pull the Drawer First: What Dovetails Prove in Ninety Seconds

A drawer is the most honest part of any case piece, because the joints that hold its corners together were cut with whatever technology the maker had on hand. Slide one out, flip it, and look at where the drawer front meets the side.

Hand-cut dovetails run from the mid-1700s through roughly the 1860s on American work, and they betray the hand every time. You will see a small number of them on a drawer, often just three or four across a side, and they will not be perfectly matched. The pins, the narrow wedge-shaped members, are usually thin and pointed on hand work, and you can frequently find the scribe line the cabinetmaker knifed across the wood to mark his cut. Look for slight over-cuts past that line, faint saw kerfs, a little slope variation from one pin to the next. Perfection is the tell you do not want if you are hoping for age.

Machine-cut dovetails arrive around the 1860s and change the look entirely. The joints become uniform, evenly spaced, and often more numerous, because a cutter head could rout a dozen matched pins in the time a person cut one. The width of the pins and tails becomes nearly identical along the whole joint. If every dovetail on a drawer is a carbon copy of its neighbor, you are looking at machine work, and the piece cannot predate the machinery.

Note which kind of dovetail you are looking at. The front corners of a drawer are almost always half-blind dovetails, cut so the joint does not show on the face of the drawer front, while the back corners are through dovetails visible from both sides. That is standard for hand and machine work alike, so the joint type does not date the piece; the cutting does. Count them, too. A wide 18th-century drawer side might carry only four or five hand-cut dovetails because each one was labor a person had to pay for in time, whereas a cutter head could afford eight or ten on the same board, so a high, tidy count leans late.

The most useful single joint for dating is the Knapp joint, and it is unmistakable once you have seen it. Charles Knapp patented it in 1867, and factories ran it from roughly 1870 to 1900. Instead of interlocking wedges, it shows a row of round pins seated into scalloped half-moon cutouts, a pattern people describe as "pin and cove" or "half-moon." No hand tool makes that shape. A Knapp joint is a firm bracket: the drawer, and therefore the piece, was built in that thirty-year window and no earlier. That is exactly what unmasked the second "Chippendale" chest at the estate sale. The form said 1770; the drawer said 1880.

The ninety-second drawer read:

• Few, slightly uneven dovetails with a visible scribe line: hand-cut, pre-1860s.
• Many identical, evenly spaced dovetails: machine-cut, 1860s onward.
• Round pins in scalloped half-moon sockets: Knapp joint, 1870 to 1900.

Reading Saw Marks on Drawer Bottoms and Case Backs

Turn the drawer over again and look at the underside of the bottom board, then check the back panel of the case and the inside faces of any wide boards nobody bothered to finish. Sawyers left signatures there too, because dressing those hidden surfaces smooth was wasted labor.

Straight, roughly parallel, slightly irregular gouges running across the grain are pit-saw or frame-saw marks, and they point to the 18th and early 19th centuries. Two men worked a pit saw, one standing above the log and one below in the pit, dragging a long blade in a reciprocating stroke. The cut wanders a little because human arms are not machines, and the marks sit at a slight angle with uneven spacing. On American furniture, this hand-sawn surface generally means the work predates the 1830s.

Curved, arcing marks are the fingerprint of the circular saw, and here the timeline needs care because it is two-tiered. Small, thin boards and decorative stock could be run through early circular blades after about 1830, and those arced marks are readable by 1860. Large structural boards are a different story: a circular blade big enough to break down furniture-width lumber did not come into common use until around 1880, so wide sawn boards showing bold sweeping arcs usually date from 1880 to 1900 and after. A drawer bottom cut in tight concentric curves is a modern surface no matter what the drawer front looks like.

Hand-plane marks add a fourth reading. A wooden jack plane leaves shallow scalloped tracks and low ridges across a board's hidden face, slightly irregular because a person pushed it down the length of the wood. A powered thickness planer leaves faint, perfectly even ripples spaced a fraction of an inch apart, the same distance every time. Run your fingertips across a backboard in raking light from a flashlight held low. Broad, uneven scallops say hand tools and pre-industrial work; a mechanical washboard evenness says a powered planer and the machine age.

Band-saw marks are the third machine pattern: fine, straight, closely spaced lines with no arc, left by a continuous blade running over two wheels. Band saws entered furniture shops after the 1860s and dominate late-Victorian and 20th-century work. If the hidden surfaces are perfectly flat with a light sanded sheen and no tool signature at all, you are almost certainly looking at machine planing from the modern era. The principle holds across categories, the same "newest evidence wins" logic that quilt historians use when they date a top by its most recent fabric and thread rather than its oldest patch (see our guide on dating antique quilts for a full framework).

Rose-Head Nails, Gimlet Screws, and the 1846 Dividing Line

Fasteners are the cheapest thing in a piece of furniture and the hardest thing to fake convincingly, which makes them prime evidence. Look at any nail or screw that appears original, meaning it sits in wood that has aged around it and shows no fresh splitting or shiny threads.

Nails run through three clear eras. Hand-forged nails, made one at a time by a blacksmith, have irregular squarish shanks and a hammered, off-center head, the classic "rose-head," and they belong to the period before the 1790s. Cut nails came next: sheared from a flat sheet of iron, they have a rectangular, tapered, wedge-like shank and a more regular head, and they carried furniture and buildings from the 1790s through the 1890s. Wire nails, the round smooth-shanked nails you buy today, were developed around 1880 and dominated the market by the 1890s. A round wire nail in a joint that has never been repaired caps the piece at the 1890s or later, full stop.

Expect some mixed hardware, and do not panic at a single modern fastener. A 1780s chest that has lived two centuries has picked up replacement screws and a repair nail or two, so one wire nail sitting in an obviously disturbed, re-glued joint is a repair, not a verdict. The test is the undisturbed joint. A wire nail in wood that has never been opened, with aged grain closed tight around the shank and the same oxidation on nail and wood, is original, and it caps the date. Judge every fastener by the wood around it.

Screws split the timeline even more sharply, and the dividing line is 1846. Early screws were made by hand, with off-center slots, uneven threads, and a blunt tip rather than a point, because nobody could machine a sharp gimlet point yet. That changed when a practical gimlet-point screw, one that tapers to a self-starting point, came out of an 1846 patent and became standard by 1860. If a screw ends in a sharp point, it was made in 1846 or later. If it ends blunt with hand-filed threads, you may be holding pre-1846 hardware.

The head style adds a second bracket. The slotted flat-head screw covers the entire 18th and 19th centuries and most of the early 20th. The Phillips-head screw, with its cross-shaped recess, was adopted on the assembly line for the 1936 Cadillac and reached general use in the late 1930s. A single original Phillips screw drops the whole piece into the late 1930s at the earliest, regardless of how convincingly Colonial the rest of it looks. When you are weighing what a period piece is actually worth once you have dated it, the pricing and buying logic in the broader furniture guide picks up where identification leaves off.

Flip It Over: Secondary Woods and What They Say About Where It Was Made

Cabinetmakers used their expensive show wood, the mahogany or walnut of the drawer front, only where it would be seen. Everywhere hidden, they reached for cheap local lumber: drawer sides and bottoms, backboards, glue blocks, dust panels. That hidden "secondary wood" is a regional postmark, because a maker used whatever grew and milled nearby.

White pine, soft and pale with a fine even grain, points to New England. Yellow or hard pine, denser, resinous, and darker in the growth rings, points to the American South. Tulip poplar, greenish or pale brown and very lightweight, is the classic Mid-Atlantic secondary wood, common on pieces from Pennsylvania, New Jersey, and Delaware, which is exactly what you would expect to find behind the drawer fronts of a chest built in Philadelphia. Oak as a secondary wood, especially riven oak used for drawer linings and backboards, is a strong flag for English rather than American manufacture. That distinction alone can move a piece across the Atlantic in its catalog description.

Regional habits get specific enough to place a piece in a city. Newport, Rhode Island, cabinetmakers in the Goddard and Townsend orbit worked mahogany over chestnut and pine secondaries; Philadelphia's high-style carvers worked mahogany over Atlantic white cedar and poplar. You will not name a maker from wood alone, but a poplar-lined Chippendale chest is telling you Mid-Atlantic before you have read a single dovetail, and that alone narrows the search.

Wood movement is its own slow clock. Boards shrink across the grain, never along it, as they dry over decades, so a genuinely old round tabletop measures slightly oval when you put a tape across it two ways, and a wide old drawer bottom will have shrunk enough to leave a gap at its edge or pull against its nails. A perfectly round two-hundred-year-old top is a warning. A top that has gone a quarter-inch out of round is doing exactly what two centuries of seasonal swelling and shrinking should do, and that is very hard to fake.

The primary wood carries a rough date map of its own, tracking the fashions that swept American shops:

• Walnut dominates the William and Mary and Queen Anne eras, roughly 1700 to 1750. American Queen Anne furniture runs about 1730 to 1760 by the Metropolitan Museum's dating.
• Mahogany takes over for the Chippendale and Federal periods, roughly 1750 to 1830, imported from the Caribbean and prized for carving.
• Rosewood, dark and dramatically striped, marks high-style early Victorian work of the 1830s through 1860s.
• Victorian walnut returns as the workhorse of Renaissance Revival and Eastlake furniture, roughly 1850 to 1880.
• Golden oak, quarter-sawn and honey-colored, floods factory furniture from about 1880 to 1920.

None of this is a lock on its own. A walnut drawer front over poplar sides with hand-cut dovetails and rose-head nails builds a consistent early story; a golden-oak dresser over pine with a Knapp joint and wire nails tells a coherent 1890s story. When the woods, the joints, and the fasteners all point to the same decade, you have identification. When they fight, the newest evidence wins.

Stickley Decals, Kittinger Brands, and Baker Tags

A signature can shortcut all of this, but marks are a blunt instrument. They are faked, swapped, and misread constantly, so a mark confirms a piece the construction already supports; it does not override construction that contradicts it. Learn the real ones anyway, because they turn a good guess into a firm attribution.

Gustav Stickley's Craftsman workshop used a red decal, a joiner's compass enclosing the motto "Als ik kan" with his signature, from 1901 to 1903, its earliest and most collectible mark. His brother's firm, L. & J.G. Stickley, ran a "Handcraft" decal built around a handscrew clamp from 1906 to 1912. The two are constantly confused, and the difference in date and maker moves real money in the Arts and Crafts market. A genuine early Gustav red decal on an honest quarter-sawn oak piece is a five-figure object; the same form unsigned is a fraction of that. Because Mission oak is heavily reproduced, treat any Stickley mark as a claim to verify against joinery, tenon-and-key construction, and wear, never as proof by itself.

Later factory makers branded and tagged their work more consistently. Kittinger of Buffalo, New York, made the licensed Colonial Williamsburg reproduction line and stamped those pieces with a "CW" prefix followed by a model number, a code that ties a given chair or lowboy to a specific Williamsburg original. Baker Furniture used metal tags and paper labels, including its crown device, on quality mid-century case goods. These marks say "good factory reproduction of a period form," which is valuable information: a Kittinger CW secretary is a well-made 20th-century piece worth real money as exactly that, not an 18th-century antique, and reading the mark correctly keeps you from overpaying for the wrong century. Once you know what you are holding, the places that will actually pay for it locally depend heavily on whether it reads as period or reproduction.

How to weigh a mark:

• A mark consistent with the construction: strong confirmation.
• A mark that contradicts the construction: distrust the mark, not the joinery.
• No mark at all: normal for most pre-1890 furniture, which was rarely signed.

When the Style Says 1770 but the Screws Say 1890: Spotting Centennial and Revival Copies

The 1876 Centennial Exhibition in Philadelphia set off a wave of nostalgia for Colonial furniture, and shops spent the next forty years building "Centennial" and Colonial Revival pieces in 18th-century forms. These are not crude fakes. They are often beautifully made, and a Centennial Chippendale highboy can be a fine object worth thousands, but it is a Victorian piece wearing an 18th-century costume, and confusing the two is how people lose money on the identification. The whole trick is letting the hidden evidence outvote the silhouette.

There is a reason the evidence hides underneath. A Centennial maker lavished skill on the show surfaces, because those sold the illusion, and spent nothing disguising the modern shop tools on the parts a buyer would never flip over. So the back of the case, the underside of the drawers, and the inside of the feet are where the truth lives. Pull the bottom drawer all the way out and look into the empty pocket at the dust board and the runners; that cavity is rarely dressed up.

Work the piece against itself. A ball-and-claw foot and a serpentine front say 1770; a Knapp drawer joint, a circular-sawn drawer bottom, wire nails, and pointed gimlet-tip screws all say 1880s or later. When style and construction disagree that sharply, the construction is telling the truth and the style is telling you what the maker wanted you to believe. Machine-perfect dovetails on a "Queen Anne" lowboy, a Phillips screw in the "Federal" backboard, band-saw marks on a "Chippendale" drawer bottom: any one of these caps the age, no matter how right the front looks.

Watch for tidier tells too. Genuine 18th-century secondary surfaces show oxidation gradients, hand-plane scallops, and shrinkage that pulled boards slightly out of round; a drawer that is perfectly square and evenly colored inside is suspiciously young. Uniform "distressing," meaning dents and wormholes that all sit at the same depth or cluster where a machine could reach, is applied age, not earned age. Real wear lands where hands and feet actually touched: the front edges of drawers, the arms of chairs, the stretchers where boots rested. Reproduction wear tends to appear in places that look good rather than places that got used. Dating a piece correctly is not a downgrade, it is a price. A Centennial mahogany highboy built by a skilled Boston shop around 1890 can bring a few thousand dollars as a fine Colonial Revival object; the same form catalogued as a genuine 1760s highboy would be a six-figure claim, and the distance between those two numbers is exactly the distance a Knapp joint or a single wire nail closes. Getting the century right protects the honest buyer and the honest seller in equal measure. Interest in these correctly-dated period pieces keeps climbing, a trend the 2026 market forecast traces across the brown-furniture category as younger buyers relearn the difference between a real Chippendale chest and a Centennial one.

Quick Reference: The Evidence Table

Keep this in your head at the estate sale. Each row is a piece of physical evidence, what it looks like, the earliest date it could exist, and the range where it is most common. Date the piece by the newest row you can find on it.

The one rule that ties it together: a piece is no older than its newest original evidence. A gorgeous hand-dovetailed Queen Anne form with a single original wire nail in an unrepaired joint is a piece made in the 1890s, not the 1750s. Find every fastener, joint, and saw mark you can, line them up on a timeline, and let the latest honest one set the date. Style opens the conversation. Construction ends it.