A patient sits down, turns toward the mirror and says "cheese," the way we all do for a photograph. Two cords tighten in her neck, running from the jaw toward the collarbone. "Can you relax those?" she asks.

Often, yes. But the question worth asking is not whether botulinum toxin can soften a neck band. It is how little it takes, and why so many people assume the answer is "more." In the neck, the evidence supports modest doses placed with purpose, because a bigger dose buys more risk than benefit.

The Bust of Nefertiti in Berlin's Neues Museum shows what patients are really asking for. The neck rises from the shoulders like a clean column and carries the face without strain. Clinicians even borrow her name for a jawline technique, the "Nefertiti lift" (Syed et al., 2026). A good neck looks light. It does not look frozen.

What the trial tested

The platysma is a thin, wide sheet of muscle that runs from the upper chest over the jawline. When it contracts, it raises vertical bands in the neck and pulls the lower face downward (Rohrich et al., 2025). That is what we are targeting with the botox.

The best controlled evidence is a phase 2 randomized, double-blind, placebo-controlled trial at 12 centres in the United States and Canada, published in Plastic and Reconstructive Surgery (Rohrich et al., 2025). The analysis covered 164 adults with moderate to severe platysma prominence. They received a low dose (26, 31 or 36 units, depending on severity), a high dose (52, 62 or 72 units), or placebo, and were followed for four months (Rohrich et al., 2025).

At day 14, investigators rated 77.8% of the low-dose group and 88.2% of the high-dose group as improved by at least one grade on both sides, compared with 12.0% on placebo (Rohrich et al., 2025). Patients' own ratings were close: 75.9%, 88.2% and 18.0% (Rohrich et al., 2025). Against placebo, the toxin clearly worked. And the photo the investigators have shows what the result can be like.

The extra units bought little and cost more

The gap between doses is where the argument lives. The high dose scored about 10 percentage points higher on the investigator scale, but the confidence intervals overlap (66.7% to 88.9% for the low dose, 79.4% to 97.1% for the high dose), and the trial was not designed to compare the two doses (Rohrich et al., 2025).

The safety data point the other way. Adverse events related to the study drug occurred in 16.7% of the high-dose group and 1.7% of the low-dose group, and in none on placebo (Rohrich et al., 2025). Five participants developed neck muscle weakness, all at the 72-unit dose, and two had mild dysphagia (difficulty swallowing) at 52 and 62 units (Rohrich et al., 2025). Most weakness was mild or moderate, but one participant had severe weakness for 10 days, then moderate for 22 days, then mild for 36 days (Rohrich et al., 2025). Even the low-dose group’s dosages seem to be quite a bit compared to what I’m used to inject in the Korean market.

The low dose was not free of trouble either. One participant at 26 units had mild weakness of a lip-depressing muscle that lasted 133 days (Rohrich et al., 2025). The authors capped the phase 3 dose at 36 units (Rohrich et al., 2025). The depression of the lip in the paper probably happened because the toxin affected the risorius muscle.

What the pooled data add, and what they hide

A 2026 systematic review and meta-analysis in the Journal of Cosmetic Dermatology pooled three randomized trials with 912 participants (Syed et al., 2026). Toxin made a one-grade improvement about four times as likely (risk ratio 4.11; 95% CI 3.60 to 4.69) and a two-grade improvement about 1.8 times as likely (risk ratio 1.83; 95% CI 1.54 to 2.17) (Syed et al., 2026). Satisfaction was higher as well (risk ratio 5.55; 95% CI 4.15 to 7.43) (Syed et al., 2026).

Some will say the safety result closes the case: adverse events were no more common on toxin than on placebo (risk ratio 0.95; 95% CI 0.76 to 1.20; Syed et al., 2026). That is partly true, and it is reassuring for lower doses. But a count of "any adverse event" is dominated by bruising, which a placebo needle causes just as readily (7.1% on placebo in the phase 2 trial; Rohrich et al., 2025). It cannot see a weak neck.

The review also has cracks. With so few trials it could not assess publication bias, its trials followed patients for 120 days after a single session, and its forest plots count the phase 2 trial's single placebo group of 53 people twice, once against each dose, so the placebo total in its figures (482) exceeds the 429 stated in its text (Syed et al., 2026). That likely makes its intervals look tighter than they are. I read its ratios as a direction, not a measurement.

Where you inject may matter as much as how much

A clinical commentary in the Journal of Cosmetic Dermatology makes a different argument. The muscle's motor nerve supply is concentrated in its upper part, and its lower third is served mostly by sensory nerves, so toxin belongs in the higher section (Yi et al., 2025). In 40 patients treated under ultrasound guidance at 15 sites per side, the authors report that 92.5% had at least a two-grade reduction in dynamic bands by week 4, maintained through week 12, with bruising in 5% and no dysphagia (Yi et al., 2025). They also report 35% less toxin than historical whole-muscle protocols (Yi et al., 2025).

This is a commentary, not a trial. There was no control group and no randomization, and the authors call their findings preliminary (Yi et al., 2025). Their dose of 30 to 60 units per side is also larger than the trial's low dose, and the paper does not name the product (Yi et al., 2025). Units are not interchangeable between brands. I read it as a hypothesis worth testing, not a protocol to copy. The study does not disclose any result pictures. The injection protocol is very different from what we see on Rohrich et al.,’s paper.

The Clinical Bottom Line

Before you book, ask four questions.

First, is the muscle really my problem? The phase 2 trial excluded people with excess skin laxity or predominant submental fat, because these can mask the bands (Rohrich et al., 2025). A sagging neck may need a different tool.

Second, what total dose, which product, and where will it go? "As much as needed" is not an answer.

Third, what should I watch for? Trouble lifting your head from a lying position, or difficulty swallowing, deserves a call to your clinic (Rohrich et al., 2025).

Fourth, what happens next? In the trial, the effect peaked at day 14 and declined afterward, so this is a maintenance decision (Rohrich et al., 2025). And usually neck botox will last anywhere from 3-4 months.

The queen in Berlin has kept her neck for more than three thousand years because it was made to carry weight, not because it was emptied of strength. Treat the band, not the whole neck.

For clinical assessments and consultation in Seoul: itsdrbock.com

References

Rohrich, R. J., Bertucci, V., Dayan, S., Jones, D., Solish, N., Rivers, J. K., Weiss, R. A., Muhn, C. Y., Harutunian, C., Park, G. S., Shimoga, S., Lee, E., & Tong, W. (2025). Efficacy and safety of onabotulinumtoxinA for the treatment of platysma prominence: A randomized phase 2 dose-ranging study. Plastic and Reconstructive Surgery, 155(1), 79–88. https://doi.org/10.1097/PRS.0000000000011472

Syed, R., Khan, A. A., Shah, S., Khan, A., Idrees, M., Ali, M., Al Sinani, M., & Al Maqbali, M. (2026). Efficacy and safety of onabotulinumtoxinA for the treatment of platysma prominence: A systematic review and meta-analysis of randomized controlled trials. Journal of Cosmetic Dermatology, 25, Article e70701. https://doi.org/10.1111/jocd.70701

Yi, K.-H., Wong, I. K. J., & Wan, J. (2025). Optimizing botulinum toxin injections in the platysma muscle based on motor nerve distribution. Journal of Cosmetic Dermatology, 24, Article e70301. https://doi.org/10.1111/jocd.70301