Crowns, only when there's no smaller answer.

A crown protects a structurally compromised tooth by enveloping it. That envelopment is genuinely useful: a tooth with a vertical crack, severe bulk loss, or extensively undermined cusps is mechanically held together better by a properly designed crown than by any partial restoration. The benefit is real. So is the cost. To make room for the crown, the dentist removes a circumferential band of enamel and dentine, typically 1.5 to 2 mm of tissue, all the way around. That tissue does not grow back.

Edelhoff and Sorensen quantified exactly how much tooth structure different preparation designs sacrifice ^[4], and the numbers are sobering, full crowns remove dramatically more tissue than partial coverage restorations on the same tooth. There is also a quieter, longer-term cost: vital teeth that are crowned have a measurable risk of pulpal complications afterwards. The Bergenholtz and Nyman study, now four decades old but still cited because the principle has not changed, documented the rate of pulp necrosis after prosthetic preparation in vital teeth ^[5]. None of this argues against crowns. It argues against crowning teeth that don't need to be crowned.

The vocabulary around the smaller options is worth knowing, because they are the ones least often named out loud. An inlay sits inside the cusps. An onlay covers one or more cusps and leaves the rest of the tooth untouched. An adhesive onlay, or bonded onlay, is the same restoration described by how it is retained: by the bond rather than by the shape cut into the tooth. A biomimetic crown, where that term is used honestly, means a restoration designed around what remains of the tooth rather than around a standard preparation form. If a plan lists a crown and none of these came up in the conversation, that is a fair thing to ask about.

When the indication for a crown is correct, modern crowns perform very well. The Pjetursson group's systematic review of single crowns reported 5-year survival rates in the range of 95% for both zirconia-ceramic and metal-ceramic restorations ^[1]. The Pieger systematic review focused specifically on lithium disilicate single crowns and reported similarly high clinical success ^[2]. The most recent meta-analysis of monolithic ceramic restorations confirmed that the move from veneered to monolithic ceramic, driven by chipping problems with veneered zirconia, has produced excellent contemporary survival data ^[3].

The survival numbers do not, however, answer the question of whether a particular tooth should have been crowned in the first place. A crown that lasts twenty years on a tooth that didn't need it is still a tooth that lost two millimetres of healthy structure. The conservative principle is the same it has always been: pick the smallest restoration that solves the actual mechanical problem.

The first question is always whether a partial restoration can do the job. If the tooth has at least one or two intact walls, a bonded onlay or large composite can often restore function with a fraction of the tissue cost. The Cochrane review of single crowns versus conventional fillings for the restoration of root-filled teeth concluded there is insufficient evidence to favour routine crowning over a well-placed conventional restoration ^[6]. That finding has been part of the conservative case against reflexive crowning for over a decade.

When a crown is the right answer, the design choices follow the same conservative logic. I default to lithium disilicate for posterior teeth where strength and aesthetics both matter; high-translucency monolithic zirconia is a reasonable second choice for very high-load situations. Veneered zirconia, with its known chipping risk, is rarely the right call when a monolithic option exists. The preparation is kept as conservative as the indication allows, the margin is placed supragingival or equigingival whenever possible, and the cementation step is treated as adhesive bonding rather than passive seating, which protects the underlying tooth in exactly the way that makes the rest of the restoration's mechanics work.