Neck Training Machines
Fixed-path loading through a pad — the highest load capacity available, and the narrowest range of what it can train.
A neck machine loads the head through a padded arm moving along a fixed arc, usually against a weight stack. It offers the highest load capacity and the most precise measurement of any category — you know exactly what you lifted — and it is the only modality shown by electromyography to reproduce the muscle activation of high-G flight. Its limitations are that it constrains the path, occupies significant floor space, and cannot train rotation at all.
How it works
A pad contacts the head — forehead, occiput, or the side — and is attached to an arm that pivots about a fixed axis, connected to a weight stack or plate loading. The user is seated, usually with a chest or shoulder brace preventing the torso from contributing.
Most units are described as four-way: flexion, extension, and lateral flexion to each side, achieved by the user turning in the seat between exercises. A minority are dedicated single- direction units.
The load profile is essentially constant through the arc, modified by the cam or lever geometry of the particular machine. Compared with a hanging weight, which loads hardest at maximum head tilt, and a band, which loads hardest at full stretch, a machine is the most uniform resistance in the category.
The one thing machines demonstrably do best
The electromyography study that established the ceiling for elastic bands used a resistance machine as its comparison, and the machine numbers are the useful part.
Measured against what pilots’ necks actually do in flight:
| Modality and intensity | Approximate equivalent |
|---|---|
| Elastic band, any grade | ~1 G |
| Machine at 50% of 3RM | 3–5 G range |
| Machine at 70% of 3RM | exceeds 3–5 G for most muscles |
| Machine at 90% of 3RM | exceeds 5 G; cervical erector spinae reached 93% MVC |
Netto KJ, Burnett AF, Coleman JL. Neck Exercises Compared to Muscle Activation During Aerial Combat Maneuvers. Aviation, Space and Environmental Medicine, 2007;78(5):478–484.
This is the only evidence in the literature that any device category reproduces a specific, externally-defined loading demand. If you need to load a neck hard and know that you have, a machine is the instrument that has been shown to do it.
The authors were correspondingly cautious about the top end: higher intensities should be “administered with care” and, in their trainee population, “limited to trainee pilots not in the high Gz phase of their training.” Ninety percent of a three-rep maximum on a neck machine is a serious load.
What it cannot do
Rotation. No conventional four-way machine trains axial rotation. The pad and arm geometry has no way to resist the head turning about its own vertical axis. Given that rotation has the tightest angular tolerance of any cervical direction and accounts for roughly a third of severe posture time in helicopter aviators, this is a significant gap that machine-only training leaves entirely unaddressed.
Combined and multi-planar movement. The path is fixed. Anything diagonal, or any movement combining directions, is outside what the machine offers.
Adapting to the individual. The pivot axis is where the manufacturer put it. If it does not correspond well to your cervical axis of rotation, the machine will load you along a path that is not quite your own, and there is no adjustment that fixes an axis mismatch.
Being anywhere convenient. These are large, heavy, expensive units. Most people encounter one in a well-equipped athletic facility or not at all — which is a real limitation given that adherence, not programming, is the binding constraint in every long-duration study on record.
Fit and setup
Seat height determines the axis. Adjust so the pivot of the machine arm is approximately level with the base of your skull. Getting this wrong is the most common setup error and it changes which cervical levels take the load.
Brace the torso. If your trunk is moving, the machine is loading your trunk. Most units provide a chest pad; use it.
Set the range stops if the machine has them. Joint reaction forces rise steeply beyond roughly 15° of flexion, 30° of extension, and 35° of rotation. A machine will happily take you past those angles under load, and unlike a harness it will do so along a path you cannot deviate from.
Start absurdly light. Machine plate increments are often 2.5 kg or 5 kg, which are enormous jumps for a neck. The best-evidenced progression in the literature advanced 0.5 kg at a time. Where the smallest available increment is too big, add repetitions instead until the jump becomes proportionate.
Where it fits
Best case: a strength programme in a facility that has one, for an athlete who needs load beyond what a harness comfortably provides, with rotation trained separately by manual resistance.
Poor case: the entire neck programme. Machine-only training means never loading rotation and never training the neck in any position other than seated and braced.
The historical line
Fixed-path neck loading is a mid-century idea rather than an early one. The earliest devices were harnesses and weighted headgear; the notion of constraining the movement while loading it appears in the patent record with devices like Bustamante’s 1957 chin-and-shoulder brace, where a crossbar pushes up under the jaw while shoulder pads press down.
The commercial four-way neck machine belongs mainly to the American football strength culture of the 1970s onward — a period that is thin in this archive because the primary sources have not yet been gathered. NOCSAE standards records and NCAA rule change history are the next place to look, and until they have been read this page’s account of that era should be treated as incomplete.
The pre-1970 record is on the patent history page.
Evidence, stated honestly
There is no trial comparing machines against harnesses, ring-and-track devices, or weighted headgear. The Netto EMG work establishes what a machine produces in terms of muscle activation; it does not establish that machine training yields better outcomes over time than any other method.
What can be said: machines can load a neck harder and more measurably than anything else, and the intensity they produce has been quantified against a real-world benchmark. That is a narrower claim than “machines are best,” and it is the one the evidence supports.