What to use on furniture, and what not to
Every material the manual mentions, what it is for, which finishes it is safe on, and what it does that cannot be undone. The last column is the one that matters in ten years.
Each row's last column is the point of this page. Plenty of products improve a surface this afternoon and leave a problem for whoever works on the piece next, and the sources below say which ones.
| Material | What it is for | Safe on | What it does that does not reverse | Who says so |
|---|---|---|---|---|
| Mineral spirits (paint thinner, Stoddard solvent) | Testing for wax and for finish solubility; the cleaning routine; removing a wax bloom | Sound, stable clear finishes that did not soften under it. Also named for cleaning where a finish is worn or cracked, in place of water | Nothing to a sound film. It removes the wax layer, so rewax when dry. A Class II combustible liquid: ventilate throughout work and drying | NPS Conserve O Gram 7/1; NIOSH Pocket Guide for the flammability |
| Denatured alcohol | Step 2 of the finish test | Testing only, on a hidden spot | Dissolves shellac. On a shellac surface it is a stripper, which is why the test uses a swab in a place nobody sees | Flexner, Popular Woodworking, April 2016 |
| Lacquer thinner | Step 3 of the finish test | Testing only, on a hidden spot | Dissolves lacquer and attacks shellac; makes a water-based finish soft and tacky | Flexner, Popular Woodworking, April 2016 |
| A weak solution of a mild soap and warm water | Stubborn dirt remaining after the mineral spirits pass | Sound stable clear finishes, wrung out well, followed by clear water and then dried completely. The leaflet names Ivory Flakes, shavings from Ivory bar soap, or Orvus paste or liquid | Water on a shellac film, on veneer or inlay used freely, or on bare wood causes its own damage. Use water very sparingly on veneered or inlaid work, one small area at a time | NPS Conserve O Gram 7/1 |
| Detergents | Nothing, here | Nothing | They leave a film that is difficult to remove and which may permanently damage some finishes. The leaflet's own exclusion names Ivory dishwashing liquid specifically, against the bar soap and flakes it allows | NPS Conserve O Gram 7/1 |
| Clean cotton cloths, turned frequently | Dusting and cleaning | Stable finishes. Wipe with the grain. A dampened cloth can reduce snagging on a stable finish; test with a dampened swab in a hidden area if water-solubility is in question | On unfinished wood, a cloth can imbed soils in the pores, and a moistened cloth can let them migrate into the wood | NPS Conserve O Gram 7/5, September 2002 |
| A vacuum with a soft brush attachment | Removing dust | Named as the most effective and least damaging method for stable finished wooden surfaces, and preferred for unfinished wood. Use a brush no wider than the nozzle and reduce suction where practical | Nothing, on a stable surface. Unstable finishes and lifting veneer should not be dusted at all: ask a conservator | NPS Conserve O Gram 7/5 |
| Feather dusters | Nothing, here | Nothing | The feathers tend to get caught in cracks and crevices and can cause detachment of fragile veneers and gilding | American Institute for Conservation, furniture leaflet |
| Paste wax, silicone-free | Protection from abrasion, moisture and airborne pollutants; a higher, more level sheen | Stable clear finishes: shellac, lacquer, varnish. Test the wax in an inconspicuous area the first time, to see whether its solvent softens the finish | Little: it is removable with mineral spirits, which is what makes it reversible. Too much makes a sticky surface that attracts dust | NPS Conserve O Gram 7/2; American Institute for Conservation |
| Paste wax on bare wood | Nothing | Nothing | It penetrates the pores and is nearly impossible to remove | NPS Conserve O Gram 7/2 |
| Wax carrying xylene or toluene | Nothing, here | Nothing | The stronger solvents can damage some finishes and are more of a health hazard. Prefer waxes carried in turpentine or mineral spirits | NPS Conserve O Gram 7/2 |
| Pigmented paste wax | Avoiding whitish specks in the pores of darker woods | Darker finished woods: the leaflet names walnut and mahogany | It deposits color in the pores. Natural or light-colored wax is suggested for lighter woods such as maple and pine | NPS Conserve O Gram 7/2 |
| Microcrystalline wax | A highly water-resistant wax | Named as suiting small wooden objects, including items combining wood and metal | Difficult to buff evenly on large surfaces, which is why it is suggested for small objects | NPS Conserve O Gram 7/2 |
| Polishes containing silicone, and most liquid and aerosol preparations | Nothing this manual recommends | Nothing | The worst entry on this page. Silicone leaves a film that is difficult to remove and interferes with future finish treatments. A new finish applied later pulls back into craters called fish eyes, and the residue can remain even when the old finish has been stripped to bare wood. Record it on your sheet as suspected if you think a piece has had it | American Institute for Conservation; NPS Conserve O Gram 7/6 |
| Furniture oils containing linseed or other drying oils | Nothing this manual recommends | Nothing | Repeated use creates a gummy, insoluble surface coating that darkens and obscures the grain of the wood | American Institute for Conservation |
| Polishes containing non-drying oils such as lemon oil | Nothing this manual recommends | Nothing | They attract and entrap dirt and grime | American Institute for Conservation |
| Titebond Original (PVA) | Regluing a joint cleaned back to bare wood | Interior work above 50 °F. Open assembly 4 to 6 minutes, total 10 to 15 minutes, at 70 °F and 50 percent relative humidity | A cured PVA joint has to be cleaned back to bare wood before it can be reglued, which is the labor in any future repair. The maker says it is not for structural or load bearing applications, and not for exterior use or where moisture is likely | Franklin International, product page |
| Titebond III (PVA) | As above | The data line says above 47 °F; the limitations text says above 45 °F, and this manual reports both. Open assembly 8 to 10 minutes, total 20 to 25 minutes, at 70 °F and 50 percent RH | As above. The maker says it is not for structural or load bearing applications, and not for continuous submersion | Franklin International, product page |
| Titebond Genuine Hide Glue | Regluing where the next repairer's ability to take the joint apart matters | Above 50 °F. Open assembly 10 minutes, total 20 to 30 minutes, at 70 °F and 50 percent RH | Less than PVA, by design: the maker says its sensitivity to moisture allows for easy disassembly of parts, which it calls a critical benefit in antique restoration. That sensitivity is also a vulnerability in a damp room | Franklin International, product page |
| Hot hide glue | Traditional joinery and repairs intended to be reversible | A glue temperature of 110 to 160 °F, with about 140 °F named best; not above 180 °F for any length of time | Reversal takes a combination of heat and moisture, and dry heat alone is said not to be enough in most cases. Hide glue is described as the only glue that bonds to itself | Old Brown Glue FAQ |
| Epoxy and cyanoacrylate | See the Limits box below | Not established by any source this manual found | Not established | No citable source found |
The Park Service's waxing leaflet names specific silicone-free paste waxes: Staples, Butchers, Johnson, Goddard's Cabinet Makers Wax, Trewax, Antiquax and Behlen Blue Label. It also names Renaissance Wax and microcrystalline wax as highly water-resistant but hard to buff evenly on large surfaces.
That list comes with the series' own wording, reproduced here because reprinting a government product list without it would misrepresent what it is: "Mention of a product, a manufacturer, or a supplier by name in this publication does not constitute an endorsement of that product or supplier by the National Park Service. Sources named are not all inclusive. It is suggested that readers also seek alternative product and vendor information in order to assess the full range of available supplies and equipment."
Two cautions of this manual's own. The leaflet was revised in 1993, so formulations may have changed and a name on a 1993 list is not evidence about what is in the tin today. And what transfers reliably is the specification rather than the brand: a paste wax, carried in a weaker solvent such as mineral spirits or turpentine rather than xylene or toluene, with no silicone.
Epoxy and cyanoacrylate. The research behind this manual found no citable guidance on when either is or is not appropriate in furniture repair, so the manual recommends neither and describes no use for either. That is a statement about what we could source rather than a judgment that they have no place; practitioners use both. What it means for you is that you will not find a method for them here, and that if somebody proposes one for your piece, the questions in Who to hire for furniture work about what will be added and what it rules out later are the right ones to ask.
Reviving a shellac finish. For the same reason, this manual names the professional technique for reworking a deteriorated shellac film and does not describe it. Cleaning furniture, waxing it, and reviving a tired finish explains why.
The glue figures. Every assembly time, temperature and clamping pressure in this table is the manufacturer's own published product data, not an independent test, and all three Titebond figures are stated at 70 °F and 50 percent relative humidity. None of those pages publishes a clamping time or a full cure time, so this manual states none. Planning a glue-up does the arithmetic and lists the same assumptions.
Marcus's four oak chairs arrived with a sticky, gummy surface from decades of aerosol polish. Reading this table, two rows apply to what is on the chairs and one applies to what he is about to do.
Mineral spirits lifted the greasy layer over three swabs, which is the first row working as described. The silicone row is why he writes "silicone suspected" on his materials line rather than nothing: the Park Service leaflet speaks of silicone being known or suspected, and the residue can survive even stripping to bare wood, so the note is for whoever refinishes these chairs in thirty years rather than for him.
And the Titebond Original row is the one he uses for the reglue, because his garage is 55 °F, which is above the maker's published minimum of 50 °F. His record names the glue, so the next person knows the joints will need cleaning back to bare wood.
Every source in the right-hand column was read for this manual on 13 November 2026, and the full entries, with URLs, are in Where the figures come from.