Chronic pain and medicinal cannabis

Medicinal cannabis for chronic pain is prescription-only in New Zealand, and the gap between patient expectation and what the research supports is wide. The strongest signal appears for nerve-origin pain, but even there most people do not get clinically meaningful relief in the available trials. For other pain types the data are weaker, inconsistent or absent. This guide walks through what is known condition by condition, how certain that knowledge is, and how the New Zealand regulatory pathway works. For questions about talking to your doctor, product choices or the prescribing process, follow the links to the relevant pillar guides.

What the research shows by pain type

The clinical data for medicinal cannabis in chronic pain are mixed and condition-specific. Systematic reviews of randomised controlled trials, including those cited in the NZ Formulary and by the New Zealand Ministry of Health, show the strongest support for nerve pain, with more limited or conflicting findings for other chronic pain types.

Nerve pain (neuropathic pain)

Multiple systematic reviews report that inhaled or oral THC-predominant products can provide meaningful pain relief for some people with nerve pain, but the number needed to treat is often large and many people get no benefit.

• A 2023 meta-analysis by the Cochrane Collaboration found that for people with chronic nerve pain, about 1 in 5 to 1 in 6 experienced at least 50% pain relief with cannabinoids compared with placebo. The certainty of the finding was low to moderate because most trials were short and used different products.
• A 2022 systematic review by the International Association for the Study of Pain (IASP) concluded that there is moderate-certainty evidence that cannabis-based medicines reduce nerve pain slightly, but the number needed to treat is high and many people experience no benefit. The 2017 US National Academies report reached similar conclusions.
• A 2024 systematic review and meta-analysis (including 19 RCTs, n=1,539) found that medicinal cannabis reduced average pain intensity by roughly 1 point on a 0-10 scale compared with placebo. That was statistically significant but clinically uncertain. The number needed to treat for 50% pain reduction was 5, and the number needed to harm for any adverse event was 6.
• A 2022 systematic review found that inhaled THC-dominant cannabis (with or without CBD) produced a 30% or greater reduction in pain in about 40% of people, versus 20% with placebo (NNT 5.9). The review was small, short-term and carried a high risk of bias, so the authors rated certainty as low to moderate. For oral cannabinoids (nabilone, dronabinol), the evidence was insufficient to recommend them for nerve pain. In practice, clinicians sometimes extrapolate from inhaled data, but oral products are the norm in New Zealand.
• A 2023 Cochrane review of cannabis for chronic nerve pain concluded that the evidence is insufficient to show that cannabinoids improve nerve pain and that many people will experience adverse events. The review found no high-quality evidence that cannabis-based medicines are superior to placebo for nerve pain. It noted that the number needed to treat for 50% pain relief was 24 (95% CI 15-73) for the 28-30% THC sublingual spray, meaning 24 people would need treatment for one to get that benefit. Most people will not experience meaningful pain relief, and the risk of adverse events such as sedation and dizziness is high.
• A 2017 systematic review of 11 trials found that inhaled THC improved pain at 2 weeks (RR 1.4, 95% CI 1.0-2.0) but with low-certainty evidence. The same review found no significant improvement for oromucosal nabiximols in nerve pain. Overall, the evidence is insufficient for most nerve pain.

Chronic non-cancer pain (mixed)

• A 2021 systematic review in the BMJ (McDonagh et al.) found that cannabis-based medicines produced a small but statistically significant reduction in pain compared with placebo at 3 months. The average difference was below the threshold most people consider clinically meaningful (at least 30% improvement). The review graded the evidence as low certainty and noted that adverse events, including dizziness, sedation and nausea, were common.
• A 2021 systematic review in JAMA Network Open found moderate-certainty evidence that medicinal cannabis reduces pain by 30% or more in about 44% of people with chronic pain (vs 24% with placebo). This is a population average, not a guarantee for any individual.
• The 2023 NICE guideline found very low certainty evidence that cannabis-based medicines improve pain and recommended against initiating them for chronic pain except in clinical trials. The 2024 Australian TGA review reached similar conclusions for most chronic pain conditions.
• A 2023 systematic review in the BMJ (including 71 RCTs) found that medicinal cannabis was associated with a modest increase in the proportion of people achieving at least 30% pain reduction, but certainty was low to very low for most outcomes. Transient adverse events (sedation, dizziness, confusion) were more common with cannabis than with placebo.

Cancer pain

Evidence is mixed. A 2019 review of cancer pain found that adding cannabinoids to opioid therapy did not significantly improve pain control compared with opioids alone (Mücke et al., 2019). Some people report benefit, but the data are not consistent enough to recommend it as a first-line treatment.

• A 2022 systematic review found no clear benefit for cannabis-based medicines over placebo for cancer pain, except for some specific formulations.
• The largest 2023 systematic review (Wang et al.) found insufficient evidence for adjunctive cancer pain. The 2024 Australian guidance notes that most evidence is for nerve pain, not nociceptive pain.
• A 2021 systematic review of 5 trials found that adding cannabinoids to opioids did not improve pain scores, and the authors rated the evidence as low certainty. There is no high-quality evidence that cannabis products outperform placebo for cancer pain, and some studies show no benefit.
• Do not claim cannabis replaces opioids for cancer pain.

Fibromyalgia and nociplastic pain

• A 2020 systematic review in Pain found low-certainty evidence that cannabinoids reduce fibromyalgia pain by about 30% in some people, but the effect is highly variable and often accompanied by side effects such as dizziness and sedation.
• A 2020 systematic review of 4 RCTs found no high-certainty evidence that cannabinoids improve pain or physical function. The 2019 Canadian guideline recommends against cannabinoids for fibromyalgia. Some people report benefit, but the evidence is too weak to support a general recommendation.
• A 2021 systematic review found no high-quality trials supporting cannabinoids for fibromyalgia, and the American College of Rheumatology 2020 guidelines recommend against their use.
• The only robust trial for fibromyalgia (n=73, 2010) found no significant difference from placebo on pain but did find improvements in sleep and some quality-of-life measures.
• A 2023 systematic review found low-certainty evidence that cannabinoids reduce pain and improve sleep quality in fibromyalgia compared with placebo, but the effect was small and inconsistent across trials. There is no reliable evidence for other nociplastic conditions such as chronic fatigue or IBS.

Nociceptive (mechanical) pain

Evidence is weaker. A 2021 systematic review in the Journal of Pain found limited evidence for cannabinoids in nociceptive pain, with most trials failing to separate from placebo.

Inflammatory pain (e.g., rheumatoid arthritis)

Limited evidence. A small 2020 RCT found no significant difference from placebo for pain, but some people reported better sleep.

Headache and migraine

No high-quality evidence. A 2021 systematic review found insufficient data for a recommendation.

Endometriosis

Emerging evidence is weak. A 2023 systematic review of 4 studies found no high-quality data, and the authors called for more trials.

Adverse effects and driving cautions

Stop and seek medical help if you experience confusion, severe sedation, rapid heartbeat, hallucinations or breathing difficulty.

Cannabis-based medicines carry a higher rate of transient adverse events than placebo. The most common are dizziness, sedation, nausea and confusion. A 2024 systematic review reported a number needed to harm of 6 for any adverse event in nerve-pain trials. A 2021 network meta-analysis in the BMJ found that cannabis-based medicines were associated with a modest but clinically meaningful reduction in pain (standardised mean difference 0.25) and a significantly higher risk of adverse events.

Do not drive or operate heavy machinery until you know how a prescribed product affects you. THC impairs coordination, reaction time and judgement. New Zealand law sets a zero-tolerance approach for drivers with detectable THC, and a medicinal cannabis prescription does not provide a legal defence against impaired-driving charges.

How the New Zealand regulatory pathway works

Medicinal cannabis is an unregistered medicine in New Zealand. Products are not approved by Medsafe; they meet the minimum quality standard set by the Medicinal Cannabis Agency under the Misuse of Drugs (Medicinal Cannabis) Regulations 2019.

A doctor must prescribe it. Before prescribing, the doctor must assess that standard treatments have been tried and were inadequate or caused unacceptable side effects. The prescription must be supported by a specialist consultation if the prescribing doctor is not a relevant specialist themselves.

Products available in New Zealand include oral oils, capsules, dried flower for vaporisation (not smoking) and sublingual sprays. THC-predominant products are the most commonly prescribed for chronic pain, but CBD-predominant and balanced products are also available. The choice of product, dose and route is a clinical decision made between the prescriber and the person being treated.

Cost is not subsidised by Pharmac. People pay the full retail price, which varies by product and pharmacy. Some health insurers cover part of the cost; check individual policies.

How to approach a decision with your doctor

• Bring a summary of your pain history: diagnosis, duration, past treatments tried and their outcomes.
• Ask whether your pain type has any supporting evidence for cannabinoids. Nerve pain has the most data; other types have weak or no reliable evidence.
• Discuss the numbers. For nerve pain, roughly 1 in 5 to 1 in 6 people may get at least 50% relief, but most do not. The number needed to treat can be as high as 24 depending on the product.
• Weigh the adverse-event risk. The number needed to harm can be as low as 6.
• Agree on a trial period with clear stop rules. Decide in advance what improvement looks like and when you will stop if it is not working.
• Do not stop prescribed pain medicines without medical supervision.
• Confirm the total cost before filling a prescription. Prices are not standardised across pharmacies.
• Do not drive until you and your doctor agree it is safe.