3D printing hit milling costs. In-office fabrication just became inevitable.

One practice paid $190 a unit to an outside lab, then made the same restoration in its own operatory for about $15 in materials, and did that 517 times in a single year. The gap was $90,455 in lab fees that never left the building, and $133,465 across two years, according to the practitioner case data published by Dental Economics. A crown that costs $15 in resin or ceramic and $190 on an invoice stopped being a clinical decision a while ago. It's a procurement decision, and most owners are still outsourcing it to a sales rep.

Here's the sentence we expect argument on: the chairside mill you're being quoted is the wrong first purchase for most general practices, and the correct move is a printer, a lab switch, and no mill at all until you're doing north of 19 single units a month. In-office fabrication is inevitable. The specific machine everyone starts with is not.

The $175 Spread Is Real, But It Doesn't Pay for Itself Automatically

Start with the number that makes the whole category interesting. The per-unit spread between outsourced and in-house crown fabrication in the Dental Economics case was $175, and on dentures it's wider still. An outsourced digital denture runs $95 to $115 a unit against roughly $500 for a traditional lab arch, a 70% to 80% reduction, and the material cost of printing that arch yourself is $20 to $40, per AvaDent's published pricing analysis.

Those spreads are what vendors put on the first slide. What they leave off is that a spread is not a profit. A spread becomes profit only after it clears financing, service contracts, resin waste, post-processing labor, redo rates, and the hours your team spends learning to nest, wash, and cure without ruining a case at 4:40 on a Friday. The equipment cost is exactly what the peer-reviewed literature keeps excluding, and exactly what kills the return in a practice that doesn't have the volume to spread it.

What Most Owners Get Wrong: Underwriting the Machine Against Crowns

The mistake is specific and it's made by good operators: they justify the capital purchase with crown lab fees, because crowns are the procedure they can count and the one the rep models. In the survey of dentists' perception of chairside CAD/CAM published in Healthcare, crowns were the top application for 57% of users, and 27.2% of respondents already had a chairside system. It's the anchor use case.

It's also the worst one to underwrite on. Single-unit crown volume in a solo general practice is usually 12 to 20 a month, not 40, and at that volume the fixed cost of a mill eats the entire spread. Meanwhile the units that come off a printer, models, surgical guides, temporaries, night guards, try-ins, denture bases, pile up at three to five times the crown count and never show up in the pro forma because nobody bills them as a line item.

The second version of the same error: counting savings per unit instead of per chair hour. Lab fee reduction is a cost line. Visit elimination is a capacity line. They are not the same size, and in most practices the capacity line is bigger.

A 180-Crown Practice Loses $7,380 a Year on the Mill

Run it. Say you place 15 single units a month, 180 a year, which is a healthy general practice, not a struggling one. At $190 outsourced, that's $34,200 a year in crown lab fees. Fabricate them chairside at $15 a unit in materials and you spend $2,700. The gross spread is $31,500.

Now the costs the spread has to clear. A mill, scanner, oven, and software bundle financed at $135,000 over 60 months runs roughly $2,600 a month, or $31,200 a year. Service and software subscription, call it $6,000. Add about 20 minutes of assistant time per unit for design, milling supervision, sintering, and finishing: 180 units is 3,600 minutes, which is 60 hours, and at $28 an hour fully loaded that's $1,680.

Net result: $31,500 minus $31,200 minus $6,000 minus $1,680 leaves you $7,380 in the hole. You bought a $135,000 asset to lose money slightly faster than the lab was charging you.

The breakeven is worth memorizing. Contribution per unit is $190 minus $15 in materials minus $9.33 of assistant time, or $165.67. Fixed cost is $37,200 a year. Divide and you need 225 crowns a year, about 19 a month, before the machine has paid for its own existence. Below that line, every unit is a rounding error against the note. Above it, the economics turn hard and fast: at 480 crowns a year the same math throws off roughly $42,000.

Our view: if you're not already doing 19 single units a month with a full schedule and a stable team, the mill is a prestige purchase wearing a spreadsheet.

Dentures Are Where the Money Actually Moved

Here's what the crown math obscures. Take a practice delivering 4 arches a month, 48 a year. At a traditional lab's roughly $500 an arch, that's $24,000. Move those same cases to an outsourced digital lab at $105 and you spend $5,040, a savings of $18,960 that costs you a phone call and a scanner you probably already own.

Print them yourself at $30 of resin and you spend $1,440, which is another $3,600 of savings. So of the $22,560 total swing, 84% of it arrives before you buy any hardware at all. On material savings alone, an $18,000 printer, wash, and cure setup takes five years to pay back. AvaDent puts the breakeven range at 73 to 534 cases depending on equipment tier, which is a polite way of saying the entry-level printer pays back fast and the fully loaded suite may not pay back at all.

Then look at the line nobody finances against. Each digital denture case eliminates 2 to 3 patient visits against 4 to 5 for a conventional workflow. At 48 cases, two saved visits each is 96 appointments, and at 45 minutes apiece that's 4,320 minutes, or 72 chair hours. Put your own production-per-hour figure in here; if your operatory produces $350 an hour and you refill only 60% of the recovered time, that's $15,120.

That number dwarfs the resin savings by more than four to one, and it is the entire argument for in-office fabrication. You're not buying a printer. You're buying back the chair time you've been giving away two visits at a time.

The Denture Study That Should Slow You Down

Now the counterpoint, because the honest version of this argument has one. A 2026 systematic review and meta-analysis in the Journal of Prosthodontics pooled the evidence on digital versus conventional complete dentures and found no statistically significant difference in total treatment cost: a mean difference of negative $357.76 in 2025 dollars, with a confidence interval running from negative $1,340.10 to positive $624.60, p equals 0.2577. Lab costs alone came in at negative $239.77, also not significant at p equals 0.1063. And the analysis excluded upfront equipment cost entirely.

Read that carefully before you argue with it. It doesn't say digital is more expensive. It says the cost advantage is not proven across the literature for this restoration type, that the spread between studies is enormous, and that the machine you'd have to buy wasn't even in the model. Anyone selling you "digital saves money, it's proven" is overstating a result that failed significance.

Our reading: the published cost research measures the wrong variable. It compares treatment cost, which is a clinical accounting question, when the operator question is throughput per chair per year. The visit count is the finding that matters, and on visits the direction is not ambiguous.

Your Lab Is Losing 2,100 Technicians by 2035

The supply side is the part almost no owner is modeling, and it's the reason "wait and see" has an expiration date. The Bureau of Labor Statistics projects dental laboratory technician employment to fall 6% from 2025 to 2035, from about 34,400 jobs to about 32,300, and attributes the decline directly to "3D printing and other laborsaving technologies" being increasingly used to produce dental parts and appliances. Median pay was $49,610 in May 2025.

That's 2,100 technicians leaving the trade. The labs that survive will be the ones that industrialized, and they'll price printed work like a commodity. The handwork, custom shade characterization, complex full-arch, the cases that need a human with a brush, gets scarcer and more expensive, because fewer people will know how to do it and the ones who do will have pricing power.

So the lab market splits in two. Commodity digital work goes to $95 a unit and keeps falling. Artisan work goes up. If your practice is sending both categories to one lab on one fee schedule today, you are currently overpaying on one half and about to be repriced on the other. That's the structural change, and it happens whether or not you ever buy a machine.

Buy the Printer, Rent the Crowns

Our call, stated plainly. Buy the printer first: it's the cheapest entry, it serves the highest unit count, and it's the one asset that pays for itself on models, guides, and temporaries even if your denture volume never grows. Move every commodity case to a digital lab this quarter, because 84% of the denture savings needs no capital at all. Hold on the mill until your single-unit count clears 19 a month for two consecutive quarters, and then buy it on the capacity argument, not the lab-fee argument.

And treat the adoption survey with appropriate skepticism. The Healthcare survey found 57% of dentists without a chairside system intend to acquire one. Intention numbers overstate purchases, always. But even half of that becomes majority adoption, and once the majority can make a crown for $15 in materials, the $190 invoice becomes very hard to defend to anyone who asks, including the payers who negotiate your fee schedule.

What We'd Do Before the Quarter Closes

Pull your last 12 months of lab invoices and split them into two piles: commodity units a printer or a digital lab could produce, and artisan work that needs a technician's hands. Get a digital lab quote on the commodity pile this week and compare it line by line against what you paid, because that switch is free money and takes one phone call. Count your single units for the last two quarters, and if the monthly average is under 19, close the mill conversation until it isn't. Then price an entry-tier printer, wash, and cure bundle against your model and temporary volume alone, and see if it clears on that basis before you credit it with a single denture.

We're watching two things that would change this view. If commodity lab pricing drops below $75 a unit as capacity consolidates, the case for owning anything weakens considerably and renting the whole workflow becomes correct. If a peer-reviewed analysis finally measures chair hours recovered rather than treatment cost, we may find the capacity argument is even larger than the $15,120 sketch above, in which case the mill breakeven moves well below 19 units and we'd owe the early adopters an apology.

Lab fees are an overhead line, and overhead is where this decision actually gets settled. If you want to see what your lab spend looks like as a percentage of collections against practices your size, and what a $19,000 swing would do to your net, run your own numbers through our free overhead calculator before you sign anything with a 60-month term on it.