About once a week, for a course of roughly six, is the standard answer, and it is a convention rather than a finding.
Here is what the published studies actually used:
Notice what is missing from that list. None of these studies was testing the interval. Each one picked a schedule, used it, and reported what happened. Nobody has run the study where one group is treated weekly, another every two weeks and another monthly, and the results compared. So "once a week" is what researchers and clinics happen to do, and it is a sensible schedule, but it has not been shown to be the best one.
Because the barrier recovers in about two days, and a week is comfortably longer than that.
The skin's barrier is the outermost dead layer plus the fats between its cells. When it is disturbed, water escapes faster than normal, and that loss can be measured. Two studies give the numbers.
In 28 people, one side of the face had diamond-tip microdermabrasion and the other was left untreated. Water loss rose immediately and at 24 hours, and was back to baseline by day 2. Skin hydration and redness rose immediately and were back to baseline by day 1. The authors wrote that weekly diamond microdermabrasion, as it is presently done, is expected to allow enough time for the damaged skin to recover its barrier function in most parts of the face. That sentence is the origin of a lot of clinic scheduling.
In 8 people, crystal microdermabrasion with two different crystal types was compared side to side. Both raised water loss at 24 hours, and by day 7 water loss had fallen below where it started.
So the interval is designed around recovery. It says nothing about whether treating weekly produces a better looking result than treating monthly, because that was never measured.
Three separate published findings point the same way.
The dose study. A randomized, investigator-blinded split-face study of 10 people ran six sessions two weeks apart, using two passes on one side and three on the other. Oil production fell from baseline on both sides, with no significant difference between two and three passes. At the first follow-up, the three-pass side actually had significantly more oil than the two-pass side. More work did not buy more result.
The adaptation study. Eleven healthy volunteers had crystal microdermabrasion once a week for five weeks, with the skin's ceramide levels measured after each session. Ceramides, which are part of the barrier's fat content, rose significantly after sessions 1 and 2, then returned to baseline after sessions 3 and 4. The skin's response faded as the treatments kept coming. That is worth holding onto when someone recommends a twelve-session package.
The depth study. In laboratory work on skin from macaques and human volunteers, the effect of microdermabrasion moved along a spectrum as passes increased, from minimal effects to extensive damage to deeper layers of the skin. At moderate settings, the outer dead layer could be removed selectively with little damage below. Holding the tip in one spot did not give clean exfoliation at all; it produced micro-blisters. Depth is not a property of the machine. It is a property of how the operator uses it.
A related finding, from work on excised pig skin rather than on patients: what controls how much surface comes off is the crystal flow rate and how long the tip dwells, not the suction pressure. Suction is the setting clinics discuss most and it mattered least.
It looks like skin that is getting worse the more you do to it, and it is a barrier problem rather than an allergy.
The outer dead layer is thin — in the one imaging study that measured it in living facial skin, the average before treatment in 8 volunteers was 9.42 micrometers, about a tenth the width of a human hair — and it is what keeps water in and irritants out. It rebuilds itself within a day or two after a normal treatment. Push harder or faster than that and you spend your time treating damage you caused.
The signs, in rough order of appearance:
Two FDA injury reports illustrate the more serious end, both from devices used at home. In one, a person reported using a home dermabrasion kit and being put on steroids to keep the inflammation in check, describing their face as beet red bordering on purple in places and being unable to be in the sun. In another, a home diamond microdermabrasion device left several bleeding scratches on first use.
The fix for over-exfoliation is not a gentler treatment sooner. It is stopping, using a bland moisturizer and sunscreen, and giving the barrier weeks rather than days.
Shorter than the schedule implies, which is part of why courses exist.
In 16 healthy women treated with a crystal device on the whole face, oil on the skin dropped immediately and was back to baseline within one hour. The immediate smoothness and matte look people notice after a treatment is largely a surface effect on that scale.
The measured structural picture is also modest. In 49 people given a single treatment, no change in the thickness of the outer dead layer was detected at all, although the skin did switch on inflammatory signals and collagen-remodeling enzymes, and 2 of 11 subjects tested showed more type I procollagen at day 14. In 38 people given eight weekly sessions, the thickness of the epidermis measured on biopsy did not change significantly in any group.
So the treatment's benefit, where it exists, is more likely a repeated low-grade healing signal than the removal of tissue. That is a real mechanism, and it is also why a single session does not do much and why a course spaced over weeks makes more sense than either a single heavy session or endless light ones.
Six is the most common published protocol. It was never derived from a dose-finding study, and nothing in the literature supports going far past it.
Clinic copy commonly recommends a course of 6 to 12. The lower half of that range matches what has been studied. The upper half does not, and it sits awkwardly next to the finding that the skin's barrier response had faded by session 3 or 4.
A practical way to run it: agree a fixed course, take a photograph in consistent light before the first session, and judge it a couple of weeks after the last one. If nothing has changed after six weekly sessions, more sessions of the same thing are unlikely to be the answer, and it is worth asking a dermatologist about a different approach.
Monthly maintenance for the indefinite future is a clinic-originated recommendation. It is usually justified by pointing at the roughly 28-day skin turnover cycle, which is a plausible-sounding rationale rather than a finding. No study has tested a maintenance schedule, measured what happens when treatment stops, or compared people on maintenance with people who stopped.
Push the interval out or stop entirely if any of the following are true.
Also space it around events rather than into them. Redness settles by about day 1 after a diamond treatment, but that is an average, and the day of a wedding is not the day to find out you are an exception.
Nobody at the FDA has reviewed a recommended treatment frequency for these machines, because nobody at the FDA has reviewed these machines.
Microdermabrasion devices are Class I, product code GFE, 21 CFR 878.4820, and are exempt from premarket review subject to the limitations in 21 CFR 878.9. Manual devices were exempted in 1994 and powered ones in 1998. Only two 510(k)s have ever been granted in that product code — K905046 in 1990 and K963204 in 1996, both before the exemption — and no machine now sold as a microdermabrasion device holds a K-number or an FDA-cleared indication.
That does not mean they are unregulated or unsafe. General controls still apply: the manufacturer must register the establishment, list the device, follow manufacturing quality rules, report certain adverse events, and not label the product falsely. What it means practically is that any frequency recommendation you are given comes from a clinic, a manufacturer's manual or a habit, and not from a regulator or a trial. It is also worth knowing that the FDA publishes consumer safety guidance for microneedling and has no equivalent page for microdermabrasion.
Infection after a device treatment is rare but real. There is a published case of a woman who developed a tender facial lump after using an at-home microdermabrasion device; it was caused by a bacterium called Mycobacterium abscessus, and clearing a single lesion took more than two months of intravenous antibiotics.