Antique furniture restoration, before the refinish.
Read the piece before you do anything to it.
n older piece is not a clean sheet of wood. It is a record of the maker’s hand, the room it lived in, and the repairs the last owner already paid for. The first thirty seconds of restoration work are not about cleaning — they are about reading. The owner who starts with a rag and a solvent has already decided what the piece is going to be; the owner who starts with a hand lens and a soft light has not yet decided, and the piece still gets to tell them what it is.
Three families of evidence, all readable without a refinish kit. Joinery: hand-cut dovetails on a drawer corner (slightly uneven, slightly proud, the saw cut visible in the cheek) are pre-machine era, almost always nineteenth century; tight, identical, machine-cut dovetails are post-1860 and the line is sharper than most owners expect. Pegged mortise-and-tenon joints on a chair or a table — a small wooden peg driven through the tenon into the mortise — are a maker's signature, not a repair. Cut nails (hand-forged, square-section, often with an asymmetric head) versus wire nails (round, uniform, bright) versus screws: the cut nails pre-date 1890; the screws tell their own story through the slot. Phillips-head screws did not exist before 1935; a slot-and-screw combination on a piece older than that is either original or a careful restoration; a Phillips-head anything in a piece from 1900 means a past repair, and a repair is information, not damage.
Finish: shellac (the dominant furniture finish in America until roughly the 1920s) feels waxy under the thumb, dissolves in denatured alcohol in seconds, and sits as a thin, slightly warm film over the wood. Early oil varnishes (copal, linseed-based) sit deeper and feel harder; post-1900 lacquers and conversion varnishes sit on top and feel plasticky to the touch. The owner who test-a small hidden corner with a cotton swab and a drop of denatured alcohol can read shellac off in under a minute; if the surface turns sticky within ten seconds, it is shellac, and the entire care plan changes around that fact. Patina is not dirt — patina is the finish thinning where the wood was handled, the color deepening where the sun never reached it, the maker’s hand visible in the places the wood was tooled and never re-finished. A piece with patina has evidence; a piece stripped of patina has a clean surface and nothing else.
Hardware: cast brass that has aged to a soft brown, hand-forged hinges with slightly uneven knuckles, escutcheons with hand-cut keyholes, and screw slots that do not line up across a single piece (because nobody used a centering punch in 1820) — all of these are signs the hardware is original to the piece or early-replaced from the same workshop tradition. Reproduction hardware is uniform in a way that the eye registers without knowing why. Drawer bottoms in secondary wood (poplar, pine, sometimes chestnut) — the wood the maker had left for a hidden surface — are a reliable dated signal on a case piece: a chest of drawers with a pine bottom and dovetailed case sides was almost always made before the maker was buying chest-wood by the kiln-dried bundle.
Patina is the evidence the piece lived in rooms with people in them, not the dirt the piece picked up in a barn. Read it before you take it off.
Restoration you can do in your own shop.
he safe DIY restoration is the smallest possible move that returns the piece to use. Not the move that returns it to "new." The owner who keeps the goal small will almost always do the right work; the owner who keeps the goal large will trigger every pitfall in the next section. The right DIY moves are the ones a competent woodworker learned in their first year — and a competent woodworker learned them by doing them on pieces they did not love, not on pieces they could not replace.
Cleaning: gentle, solvent-aware, never abrasive. On shellac, a soft cloth barely damp with denatured alcohol (a teaspoon at a time, not a soak) lifts grime off the surface without lifting the finish. On early oil varnishes, the same approach with a touch of the right wax as the lubricant. On post-1900 lacquers, a damp cloth with a drop of pH-neutral dish soap; on a piece whose finish you cannot identify, the rule is the same: the gentler answer is the right answer. Never an abrasive pad. Never a scouring powder. Never a "restorer" polish sold in a tin with a picture of an antique on the label — almost every one of those products contains a mild abrasive that removes the patina the owner is trying to keep.
Structural repair: re-gluing a loose joint is the most-rewarding DIY restoration an owner can do, and it is the one a woodworker with a decent clamp kit can do well. Disassemble the joint, scrape the old glue (hide glue residue if the piece is pre-1940 — most pre-1940 furniture was assembled with hot hide glue, and that is the glue to use again), re-fit, re-glue, clamp, let it cure. Sistering a split along the grain (a thin cleat of the same species, glued and screwed from the underside where the screw will not be seen) is the second-most-rewarding, on a tabletop or a chair seat. Replacing a missing veneer patch with a piece of the same species, cut slightly oversized and trimmed flush after the glue cures, is the third. All three stay clear of the finish; all three leave the patina intact; all three are reversible if a future restorer needs to go further.
Replacement parts from a maker: lathe-turned parts on a chair or a table — a foot, a stretcher, a finial — are the most-commonly-broken component on a piece that has lived a long life. The owner who needs one should find a turner who can match the species and the turning, not a parts catalog. Catalog parts are close enough to be tempting and wrong enough to be visible; they read on the piece as "this is not from here." A hand-cut turning from the same wood, in the same workshop tradition, with the same tooling marks, sits into the piece and disappears; a catalog part sits into the piece and announces itself.
The four refinishing pitfalls.
efinishing is the move that most often destroys the evidence the piece was carrying. It is also the most common restoration an amateur attempts, because it is the most visibly satisfying — the piece "looks new" by the end of the weekend. The four pitfalls below are where the weekend goes wrong, and they are pitfalls whether the approach is a stripper, a sander, a stain, or a re-varnish.
Chemical stripper: most strippers sold for furniture are either methylene-chloride based (fast, harsh, dangerous to the owner, and prone to lifting the patina right along with the finish) or soy-based (gentler, but slower and given to leaving a residue that the next finish will not bond to). Either way, the stripper does not discriminate between the original finish and the patina the original finish was building. A piece that came out of a strip bath is a piece that has lost a layer of its life, and the years of careful hand-waxing that built that layer are gone in an afternoon.
Sandpaper grit progression: a refinisher who starts at 80 grit and walks the piece down to 220 has erased the maker's tool marks. The plane tracks in the top of a chest, the chisel marks in a drawer face, the slight unevenness in a dovetail cheek — all of those are evidence of the maker's hand, and all of them are the first thing the sandpaper takes off. The right lowest grit on an older piece is 180, and the right highest is 220; any "refinisher" who reaches for a 60 or an 80 has decided the patina is in the way.
Over-staining to "match" a patch: the instinct to stain a replacement patch to match the aged wood around it is the wrong instinct. The aged wood has aged because of its position in the room, the light it took, the wax the last owner applied; the patch has not been there. The right move is to leave the patch a shade lighter than its surround, accept the visible patch as evidence of a careful repair, and let the patch catch up over the next ten years. A piece that has been stained to "match" reads as a piece with a lie on it; a piece with a learning patch reads as a piece that was repaired by someone who understood what a repair is.
The lost-century trap: the worst-case refinish is the one that takes a 1780 chest back to bare wood, re-stains it to a "richer" walnut the 1780 maker never used, re-finishes it with a 1960 lacquer that no period finisher would have recognized, and ends with a piece that has lost its century. The owner looking at it cannot tell what year it is from; the future owner cannot sell it as a period piece; the maker (if they were reachable, which they are not) would not recognize it as their own work. The lost-century trap is refinishing that solves the owner's discomfort with the surface and creates a piece that nobody — not the owner, not the future owner, not the restorer the next owner eventually calls — can read.
A refinish resets the piece to year zero and writes a year over the maker's hand in the wood. The hand does not come back. The future owner cannot read what was under it.
The DIY vs. hire-a-restorer line.
here is a line. The owner should learn where it is before the project starts, not after the stripper is on the piece. On the safe side of the line: any cleaning that does not involve a solvent beyond denatured alcohol; any re-glue of a joint the owner can disassemble without forcing; any patch in veneer with a piece of the same species; any wax refresh with a wax the piece has tolerated before. On the side that requires a restorer: veneer inlay (matching grain, fitting a complex shape, working the patch in without a visible seam); caning (re-caning a chair seat is a two-day project requiring a specific tool kit and a specific spline stock); gilding (every gilded surface is a finish in itself, and gilding over a damaged gild is a restorer's job); closed-grain marquetry (anything where a slip of the cutting tool costs the piece a piece of itself).
The more useful rule: if the finish on the piece predates the restorer's working lifetime as a craftsperson, the restorer should be the one who decides what to do. A restorer whose practice covers pre-1900 work has seen a thousand shellac finishes and a thousand hide-glue joints; a restorer whose training ended in the catalyzed-finish era has not, and the piece will be over-cleaned, over-sanded, or re-finished with the wrong product. The right restorer for an older piece is the right restorer for what the piece actually is — and the owner who picks that restorer is the owner who gets the right answer back.
A care record across decades.
very repair a piece has earned — every re-glue, every patch, every re-upholstery, every finish touch — is part of the piece the next owner inherits. The piece that arrives at its third owner with no record of its first two lives is a piece the third owner has to read from scratch, and the third owner will almost certainly do less good work on the third repair than the second owner did on the second. The piece that arrives with a record is a piece the third owner can keep reading; the third repair sits on top of the second, sits on top of the first, and the piece's history deepens instead of resetting.
Makerstead is the quiet ledger that rides with the piece. A QR sticker on the underside, a permanent record keyed to the maker's work order: the wood species, the secondary woods, the joinery, the finish, the hardware, the wax and the soap, the maker's contact. The second owner scans the code, sees what the piece is, and inherits the cadence the first owner built. The third owner scans and inherits the second owner's repair along with the first owner's. The piece survives into its next generation of use almost intact — not because nobody ever refinished it, but because the ledger kept the chain readable every time someone did. The maker stays reachable because the maker's contact is on the record; the restorer the second owner called stays reachable because the second owner's note is on the record; the restorer the third owner might call already has the context before they pick up the phone.
Every repair rides with the piece. The piece with a ledger outlives the ledger's first owner and reads the same to the next.
The mistakes new owners make in year one.
he most common mistakes are not refinishing mistakes — those come later, when the owner has lost patience with the patina. The most common mistakes come in the first season, when the new owner is still learning how the older piece wants to live in a modern room.
Do not place a pre-1940 case piece in direct afternoon sun; the thermal stress will check the finish inside a season and the older glue will dry out around the case joints. Do not park a piece directly over a heating register or a baseboard heater; the localized dry heat is the same stress in a smaller area. Do not cover a piece with a plastic drop cloth during a renovation; the trapped moisture will blush the original finish and the blush is not always recoverable on hide glue and shellac. Do not strip a finish to "see what is under it" — the finish is the evidence; what is under it is the wood, and the wood is not what the piece is. Do not deep-clean with a silicone polish because someone on a forum said a piece looked dull; the silicone will penetrate the finish and the piece will resist every future repair for the rest of its life.
The list is not exhaustive; the room decides. The owner who reads the piece, learns what the maker put into it, and admits when the answer to a problem is "stop and call someone who has done this a thousand times" is the owner who will still have the piece in year twenty. The owner who treats the piece as a problem to solve in a weekend is the owner who will be looking for the next piece by year ten.