Research · Manufacturers
The measurement bar just rose
A new edition of the curing-light standard and a final FDA guidance arrived within three months of each other. Read together, they say one thing: the number on the label must trace to measurement.
Two documents landed this summer that anyone who makes, sells or tests a curing light should read. In June 2026, ISO published the third edition of ISO 10650, the one standard for dental curing lights. On September 2, 2026, the US FDA issued its final guidance on what it expects in a curing-light premarket submission, two years after the draft, and replacing guidance that had stood since 2006. This article is about what they change if you make these products. The short version: both documents push toward a data-driven standard, and that means more measurement.
The ISO 10650:2026 standard is now at its third edition
ISO 10650:2026 didn't rewrite the physics. The measurement methods carry over, and so does the specification that matters most: a light is in spec when its output is within ±20% of the manufacturer's stated radiant exitance, with a ceiling of 4,000 mW/cm² (380 to 515 nm). Those are type tests, performed on representative units, not a duty to test lights in the field. What the latest edition adds is a new Type 3 class for curing lights powered from the dental unit itself (the kind integrated into the chair or delivery system rather than a standalone handpiece), and two new obligations for the instructions for use: 1. information about protective filtering devices, and 2. a version identifier on the IFU itself. I wrote about the standard in detail in the ISO 10650 explainer in August.
The regulator rewrote its playbook
The FDA guidance is the bigger shift, because of how specific it is. It's nonbinding, but in practice it describes what reviewers will expect to see in a new curing-light submission. It also states the clinical stake in its own words: inadequate radiant power output "can result in incomplete curing of dental restorative resins and lead to premature failure of the restorative material." That sentence is why Bluelight exists. By improving light curing, we help clinicians and their patients avoid pain, premature failures and unnecessary rework. For radiant output alone, it asks that a submission characterize the light "using test methods that conform to the following currently FDA-recognized consensus standard, ISO 10650," and provide six things, in the regulator's own words:
Scroll sideways to read the full list
Pause on the last two. Until recently, an irradiance-at-distance curve and a hot-and-cold-spot beam image were the concern of a handful of measurement labs. Now they're expected in a premarket file.
The guidance goes on to heat: the maximum temperature of the body and of the tip under normal and single-fault conditions, at the highest radiant exposure, after a clinically relevant curing time. One test didn't make it from draft to final. The 2024 draft asked for a depth of cure result on a legally marketed restorative resin; the final guidance does not. So the light is judged on what it emits, and the material has its own standards to meet.
The labeling recommendations complete the picture. Alongside the output numbers, the guidance recommends the instructions for use state the "recommended distance (mm) and angle (degrees) of use from the tooth surface" and give "instructions on how to periodically check the irradiance output." The idea that a light's performance should be checkable after it ships, by the people using it, is now part of what the regulator expects on the label.
Inadequate radiant power output from dental curing lights can result in incomplete curing of dental restorative resins and lead to premature failure of the restorative material.
Read them together
One document is international and voluntary; the other is American and, since September, final. Read together, they land on a single idea: a performance claim about curing is a measurement claim, and the measurement must exist. A spectrum, not just a single irradiance figure, produced on instruments whose own calibration can be defended.
For a materials manufacturer the convergence lands from the other side. With depth of cure dropped from the light's file, the material's own instructions, the increment thickness and curing time printed under ISO 4049, are where delivered energy meets cure. The two files still meet in the middle; each side now carries its half of the evidence.
What I welcome
Three things in the guidance are good news for anyone who cares about how restorations last. FDA names ISO 10650 as the consensus standard for characterizing output, so there's one method to build a lab around, not several. The ultraviolet boundary now sits at 380 nm, where the standard puts it, so a manufacturer is no longer measuring to two definitions. And the list is specific: total power, irradiance at the tip, the spectrum, dose by mode, attenuation with distance, the beam itself. Each is something a lab can measure and report the same way, every time. Add the instruction to check output periodically, and the document has, for the first time that I'm aware of, put the light's performance where it belongs: in the hands of the clinicians who use it.
None of this is new to Bluelight. We've included every one of those six measurements in curing-light assessments for manufacturers for years, alongside the heat and endurance tests, because they are the tests that tell you whether a light will do its job in a patient's mouth. What's new is that they're no longer optional. A curing light seeking US clearance, whether it's new or has been modified, is now expected to arrive with this evidence in its file, and the reviewer reading that file has a list. If you're preparing one, we can measure your light against every item on that list and help you understand where you stand before the reviewer does. The report is unbiased, because we report measurements and don't provide endorsements, and that's exactly what makes it useful to you.