Does NFC work through glass, wood and plastic?
Yes to all three. A 13.56 MHz magnetic field passes through non conductive materials without meaningful loss, so glass, wood, plastic, leather, paper and paint are effectively invisible to it. The only cost is the thickness they add between the two antennas.
What the field passes through
A magnetic field at 13.56 MHz is unaffected by materials that do not conduct electricity and are not magnetic. That covers most of what a tag is likely to be mounted behind.
Glass, ceramic, wood, plastic of every kind, leather, fabric, paper, cardboard, plaster, most paints, resin and rubber all fall into this category. A tag behind any of them works exactly as it would in the open, minus the distance the material occupies.
What stops it
Conductors. Metal in any form, including foil, metallised film, conductive ink and metal mesh. Also carbon fibre, which people rarely expect, and metallic paint, which looks like paint.
The mechanism is not absorption, it is opposition and detuning, which is why the failure is abrupt. A tag behind a sheet of aluminium foil does not read weakly, it does not read.
The one that catches people out
Metallised packaging. Crisp packets, coffee bags, blister packs, some gift wrap and many food films carry a vanishingly thin metal layer that is perfectly sufficient to stop a read while looking like plastic.
Thickness is the real budget
Since the working distance is only a few centimetres, everything between the two antennas comes out of a small budget.
A tag behind five millimetres of wood is fine. Behind three centimetres of wood, a phone will probably never reach it, not because wood blocks anything but because there is no range left.
This is why the answer to "does it work through X" is nearly always "yes, if X is thin". The material rarely matters and the thickness always does.
The surprising cases
Water and anything containing it behaves differently from dry non conductors, reducing range rather than being transparent. Wet wood is not dry wood.
Concrete varies enormously depending on moisture and on the steel reinforcement inside it. A tag behind a plastered wall may work, and a tag behind reinforced concrete will not, and the difference is the steel rather than the concrete.
Human tissue absorbs and detunes, so a tag pressed between a phone and a hand reads worse than one on a table. This is a genuine effect and it explains a number of otherwise puzzling inconsistencies in testing.
Toughened glass with a coating is worth checking rather than assuming. Ordinary glass is transparent to the field, and some coated architectural glass carries a metallic layer that is not.
Frequently asked questions
Can I put a tag behind a wall?
Behind plaster or plasterboard, yes, if it is close to the surface. Behind reinforced concrete, no, because the steel does what any metal does.
Does a screen protector affect reading?
Not the protector itself, which is glass or plastic, and the antenna is on the back of the phone anyway. If reading changed after fitting one, look at the case rather than the screen.
Will a tag work inside a sealed plastic box?
Yes, provided the wall is thin enough that the phone can still reach the tag. Plastic contributes nothing but distance.
What about tags embedded in resin or epoxy?
They work, and encapsulation is a common way to make a tag durable. The resin adds thickness, so allow for slightly reduced range compared with a bare tag.