What actually helps a stiff neck? For the sudden kind, the evidence is thin

Waking up unable to turn your head is common, alarming, and — reassuringly — usually short-lived. It is also one of the least studied complaints in musculoskeletal medicine. When Cochrane reviewers went looking for trials on how to treat acute neck pain, the sudden crick that resolves within days, they found none [s1].

The gap at the acute end

The largest review of exercise for neck pain screened 27 trials covering 2,485 analysed participants (out of 3,005 randomised) [s1]. For acute neck pain specifically, no evidence was found [s1]. That is not the same as saying exercise fails for a stiff neck; it means the question has never been tested in a randomised trial. The reason is partly that most acute neck pain settles on its own, which makes it hard to show any treatment beating the natural course. For the everyday variety, then, the honest advice is modest: it will most likely ease within days to a couple of weeks whatever you do, and keeping the neck gently moving rather than rigidly guarded is the pattern that back-and-neck guidelines endorse.

Where the evidence firms up: chronic neck pain

For neck pain that lingers, the review has more to say — though it stops short of strong claims. Moderate-quality evidence supported specific strengthening exercise. Combined cervical, shoulder and scapulothoracic strengthening and stretching produced a small-to-large benefit on pain, with a pooled standardised mean difference of −0.33 (95% CI −0.55 to −0.10) immediately after treatment, and a medium benefit on function, SMD −0.45 (95% CI −0.72 to −0.18) [s1]. Focused strength training of the neck, shoulder and upper limb produced a moderate-to-large pain benefit, SMD −0.71 (95% CI −1.33 to −0.10) [s1].

The important qualifier is that the type of exercise mattered. When trials used stretching alone, no beneficial effect could be expected [s1]. The signal was for active strengthening and endurance work targeting the neck-and-shoulder complex, not gentle range-of-motion stretches on their own. And the reviewers were explicit that no high-quality evidence was found, leaving genuine uncertainty about how well exercise works even where it helps [s1].

What about clicking the neck back into place?

Manual therapy — the mobilisation a physiotherapist does, or the high-velocity thrust a chiropractor or osteopath applies — is the other mainstay. A companion Cochrane review pooled 51 trials covering 2,920 participants [s2]. Its central finding was that the two techniques come out roughly even: for acute and chronic neck pain, multiple sessions of cervical manipulation produced similar changes in pain, function, quality of life and patient satisfaction to multiple sessions of cervical mobilisation [s2]. A single manipulation, in three small trials, relieved pain immediately but not at short-term follow-up [s2].

In other words, the more forceful thrust does not clearly outperform gentler mobilisation. Given that neck manipulation carries rare but serious risks — including injury to the arteries supplying the brain — the absence of a clear advantage is itself worth knowing. The evidence does not single out the dramatic "crack" as the thing that works.

The red flags

Most stiff necks are mechanical and benign, but some warrant urgent attention rather than exercise or manipulation. A stiff neck with fever, severe headache and sensitivity to light can signal meningitis. Neck pain after a significant fall or collision, or pain accompanied by arm weakness, numbness or pins-and-needles spreading into the hand, or by problems with balance, walking or bladder control, points to nerve or spinal-cord involvement and needs prompt medical assessment. So does neck pain with unexplained weight loss or a history of cancer. None of the trial evidence above applies to those situations.

The honest summary

For the sudden stiff neck most people mean by the phrase, there is no trial telling you what to do — because it typically clears by itself, and gentle movement is the sensible default [s1]. For neck pain that has settled in, the best-supported active treatment is targeted strengthening of the neck and shoulders, with stretching alone unlikely to do much [s1]. Manual therapy can help, but the gentle and the forceful versions perform about the same, which removes much of the case for the riskier one [s2]. Across the board the evidence is graded low to moderate, so anyone selling a definitive neck cure is going beyond what the trials support.