This is a write-up from my weekly Monday livestream, where I run a neurofeedback session on myself and answer questions live. This week's topic was sensorimotor rhythm, the 12 to 15 Hz band that sits at the center of how neurofeedback works. Audience questions are paraphrased and anonymized.
What is sensorimotor rhythm (SMR)?
SMR is a low-beta frequency, roughly 12 to 15 Hz in adults, and it shows up in one specific place: the sensorimotor strip. That is the band of cortex running ear to ear across the top of your head. Ascending sensory information from the body comes up into the sensory strip, descending motor control runs out through the motor strip, and together they produce this rhythm.
The frequency looks like beta, but it behaves like alpha. Alpha is an idle, a pulling-back. SMR is the motor system's version of that idle. When SMR is strong, the motor system is quiet and the body is still while attention stays stable. Think of a cat on a windowsill: liquid body, completely still, laser focus on the bird outside. That is a high-SMR state.
This location specificity matters. The same 13 to 15 Hz frequency measured somewhere other than the sensorimotor strip is just regular beta processing or fast alpha sliding toward beta. On the C strip, it is SMR, and it does something distinct.
Why is SMR the same thing as a sleep spindle?
SMR and the sleep spindle are the same thalamocortical circuit firing in two different states of consciousness. Awake, this rhythm holds you still and calm and focused. Asleep, the identical circuit generates sleep spindles that protect your sleep.
Here is what that looks like on a trace. A car drives by, or a dog barks two houses over. Your brain throws a large sharp waveform called the K complex, then a tight burst of fast squiggles right after it. That burst is the spindle, the same 12 to 15 Hz rhythm, and its job is to suppress wakefulness so the noise does not pull you out of sleep. The K complex got its name from sleep labs decades ago, where technicians would knock on the bed to test depth of sleep, hence "K" for knock.
That overlap explains a pattern I see constantly in the brain maps: training SMR during the day often tracks with both better daytime focus and better nighttime sleep. You are strengthening the same circuit that runs both jobs. The Hoedlmoser work showed daytime SMR training increasing sleep spindle density and improving declarative memory consolidation (Hoedlmoser et al., 2008). The spindle appears to coordinate with hippocampal ripples that move memories from short-term storage into long-term consolidation. For more on the sleep side, see SMR Neurofeedback: Train Sleep, Focus, and Self-Control and Biohacking Sleep.
How does an SMR neurofeedback session actually work?
For the stream I set up a single-channel session. One silver electrode at C4 (right sensorimotor cortex), left ear as reference, right ear as ground, running through a QWIZ amplifier into the eVox/EEGer software.
SMR is usually 12 to 15 Hz, but I trained 11.75 to 14.75 on myself. My own SMR runs a little slower these days, and it was 6pm. If I chased a true 12 Hz reward at night, I would be wide awake two hours later. Undershooting my own peak frequency slightly at the end of the day is a deliberate choice, and decisions like that are most of the skill in neurofeedback.
The setup rewards two things at once. I want my theta (4 to 7 Hz) to come down, and I want my SMR to stay up. When my brain spends half a second making less theta and more SMR, a beep fires and a block of the picture reward unveils. It is mostly involuntary. You sit, you let it happen, and the brain learns.
I also added an inhibit band in the 22 to 34 Hz range. If that fast beta gets large, it stops the reward. The reason is muscle. Muscles are electrical too, and because they sit outside the skull they are louder than the brain. Without a fast-beta inhibit you can accidentally train people to tense up to drive the game, and you can reward beta spindles that produce anxiety. For more on what the slower bands mean, see Decoding Alpha Waves.
What does neurofeedback feel like?
About one or two people out of twenty feel something on their first or second session. After three or four sessions most people feel it. It is subtle: a little calm, a little focused. There is a within-session effect that wears off in about two hours, then a quieter background effect that can run up to 24 hours and shows up as changes in sleep, attention, and stress.
By six or seven minutes into my own session I felt the shift. My resting amplitudes had not moved much, theta still high, SMR still hovering around 11 microvolts, but the feeling was there. For me it resembles the moment you stand up after an exam and set the effort down. You were not consciously doing anything, yet it feels like you just finished a cognitive task. Some people get hungry afterward, like a small workout burned blood sugar. I am one of them.
If you train a beta frequency just above where you need it, say 16 Hz when you needed 15, you get tired in the session and wired two hours later. That is overtraining, and it means the frequency choice was wrong. A good session feels low-key, and the changes accumulate over weeks.
Can you control your own brainwaves voluntarily?
Once your brain knows how to do neurofeedback, you get real voluntary control. On the stream I dropped my theta on command, from 26 down toward 18 microvolts, while keeping my SMR mostly steady. That last part matters: if everything drops together, you are just relaxing and letting muscle tension fall away. Holding the beta up while pulling theta down is the trained skill.
I trained my own theta-beta about 25 years ago, 18 sessions over five or six weeks. I went from some of the worst ADHD you have ever seen to someone who can move in and out of focus more or less on demand. That is the long-run payoff of training this circuit. For the full treatment, see Does Neurofeedback Work for ADHD?.
What does theta have to do with ADHD?
Theta is disinhibition. It takes the brakes off and lets things through. The C4 wire sits over a circuit whose job is to supervise attention, and that circuit runs on beta. When theta is high and beta is low at that location, you see the signature of ADHD and distractibility: the executive tissue cannot pump the brakes against external stimuli.
The theta-beta ratio became a marker because of Monastra's work in the 1990s, which separated ADHD and non-ADHD groups at high accuracy on EEG alone, no interview, no performance test (Monastra et al., 1999). Then something strange happened. Each attempted replication came back weaker than the last. One read on this is sleep. The adolescents in these studies were getting progressively more sleep-deprived across the decades, right alongside the rise of cell phones, and the theta-beta ratio itself has drifted across cohorts (Arns et al., 2013). The theta-beta ratio is more specific for ADHD in the absence of sleep problems, but sleep problems co-occur heavily with ADHD and with teenagers in general. The marker still indicates an executive-function pattern; it no longer cleanly indicates ADHD proper.
In the QEEG, classic ADHD shows high theta on the right (disinhibition, impulsivity) and excess alpha on the left (inattentiveness). When those bands dominate over the low beta, and you also see the inattentiveness and impulsivity on a performance test, you are in ADHD territory, whether the cause is lifelong wiring, a concussion, or chronic stress and fatigue. For the parenting angle on all of this, see Why Does My ADHD Kid Make Me Yell?.
How do you measure whether neurofeedback is working?
Two tools, run at baseline and then every other month. The first is a continuous performance test. We use the IVA-2, which measures attention and response control against age-matched norms from age six up into the eighties. Average is 100, plus or minus 15 covers about two-thirds of the population, and scores below 85 start getting in the way. The IVA-2 reads inattentiveness, impulsivity, reaction time, and stamina, and people tolerate it better than the longer alternatives.
The second is QEEG brain mapping: a cap, gel at each site, ten minutes eyes closed and ten minutes eyes open. Resting brain maps are stable month to month unless you are actively changing the brain. For what that involves, see QEEG Brain Mapping: What It Is and What to Expect.
When someone genuinely has an executive-function need, training this theta-beta and SMR work often moves performance scores by roughly a standard deviation, a 15-point swing on the bell curve, over a few months. That is a pattern I see in the maps and the performance scores, not a guaranteed outcome. On the stream I showed one client's data: 20 sessions, theta trained down along with a lot of alpha, with a clear jump in performance. The right-side theta was still there, so the next phase targeted it directly. That is the cycle: map, intervene, remap, adjust.
Typical timelines: three to four months for ADHD, trauma, anxiety, or alcohol craving; four to six months for post-concussion syndrome or seizures; longer for major developmental presentations.
Where did neurofeedback come from?
The origin is an accident. In the mid-1960s, Barry Sterman at UCLA was testing how dangerous rocket fuel vapor was for NASA, because astronauts were getting sick. He exposed cats to monomethylhydrazine and watched. Many of the cats followed a clean dose-dependent curve toward seizures.
A subset of cats resisted, holding out far longer (Sterman et al., 1969). Sterman could not explain it until he realized those same cats had been in an earlier experiment, where he had rewarded them whenever they produced SMR, conditioning the rhythm up through operant conditioning (Wyrwicka & Sterman, 1968). Months later, those cats had seizure-resistant brains.
He then built an auditory feedback device for a lab assistant with uncontrolled epilepsy, and SMR training reduced her seizures (Sterman & Friar, 1972). That is where this whole field starts. For the broader evidence picture, see Is Neurofeedback Legitimate?.
Why does training SMR produce an anti-seizure effect?
The mechanism lives in the thalamus. The sensorimotor strip carries an enormous number of descending neurons heading down through the thalamus and ascending neurons coming up from the body. Wrapped around the thalamic switchboard is a sheet of tissue called the thalamic reticular nucleus (NRT). Every corticothalamic and thalamocortical fiber passes through this NRT layer, and the NRT synapses on all of it and supplies inhibitory tone, dampening the firing.
Strengthening SMR appears to strengthen that gating. You build inhibitory tone, the brain rests with less theta, impulsivity drops on performance testing, and you raise the seizure threshold. This is also why deep pressure works: Temple Grandin's squeeze machine relieved her because deep muscle pressure triggers SMR. Inhibit the body and you create that calm, gated state in the brain. The cat on the windowsill again.
Can neurofeedback help Parkinson's, MS, or brain fog?
For early Parkinson's where the problem is rigidity, stiffness, slowed processing, and micrographia, I have seen neurofeedback track well in the maps and the reports I gather. The pattern I observe is reduced motor stiffness and, for many of these clients, dopaminergic medication that seems to work more reliably. I see younger clients in their 50s and 60s whose meds are wearing off too fast, forcing a middle-of-the-night second dose. With training, sleep often improves, the medication seems to last longer, and the night dose sometimes becomes unnecessary. A couple of people with chemically induced Parkinson's in their 40s reported their progression arrested. This is my observation as a coach, not randomized trial data, and none of it is a treatment claim.
MS is a different problem. The driver is oxidative stress, and once there is substantial tissue damage, neurofeedback has less room to change the tissue, the same limit you hit with Alzheimer's. For someone early in the MS course, the research points toward metabolic interventions: low-sugar, low-carb, higher-protein eating with moderate fat and light ketosis as an anti-inflammatory, pro-healing signal, plus B vitamins and micronutrients, saunas, ice baths, and hyperbaric work. Citicoline (CDP-choline) is interesting here too; it appears to support membrane synthesis by changing how neurons use choline in their cell membranes (Secades & Lorenzo, 2006). See Strategic Fasting and Brain Biohacking with Photobiomodulation, which I would bring in for both Parkinson's and MS to feed the brain's energy systems.
Much of brain and body aging in these conditions traces to glycation and oxidation, the rusting of tissues. Amyloid plaques oxidize, which is part of why they are so disruptive. In Lewy body disease, the edges of the Lewy bodies are glycated, and the more glycated they are, the faster they appear to spread through healthy tissue. Drop the sugar, raise protein absorption, keep good fats, and you reduce the glycation load. For post-COVID brain fog I covered on the stream, the QEEG showed excess delta with the alpha dropped out; over three months and 40 sessions, clearing the delta and restoring alpha tracked a 20-to-30-point gain on her performance scores. See Biohacking Brain Fog and The Critical Aging Window.
How do you set thresholds in neurofeedback software?
A question came up about finding correct thresholds in tools like BioExplorer. The principle comes straight from learning theory. Reward nothing and no learning happens. Reward exactly half the time and no learning happens. Reward about a third or two-thirds of the time and learning happens.
In practice: for a reward band like SMR, pass it 65 to 70% of the time. For a slow inhibit band like theta, pass it about 75% (inhibit a quarter of the time). For a fast inhibit band up in the 22 to 34 Hz range, be generous, pass it 85% and only catch the big excursions. In BioExplorer, read the greater-than and less-than signs on each threshold block carefully, because the logic is easy to invert. You can also use adaptive thresholding, but set discrete thresholds first, train a few minutes, then turn auto-thresholding on so it has somewhere sensible to start.
The bottom line on SMR
SMR is the calm-cat rhythm of the sensorimotor strip, the same circuit that becomes your sleep spindles at night. Train it up and you build inhibitory tone, which is associated with better impulse control, lower resting theta, a higher seizure threshold, and steadier sleep. It is the rhythm the whole field was built on, and it is still one of the most useful things to train.
If you want to see where your own theta, alpha, and SMR sit, a QEEG is the place to start. Peak Brain runs a summer special with $250 off a brain map, and you can book a consult through the site.