Honey for Cough and Sore Throat: What NICE Says

For a cough in a child over one, honey probably works — for the first few days. That is the finding of a Cochrane systematic review of six randomised trials in 899 children: honey probably relieves cough symptoms more than no treatment, diphenhydramine and placebo, with the advantage running to about three days. It is a modest effect with honest edges — the reviewers’ own summary sentence is that “There was no strong evidence for or against using honey” — and for sore throat specifically the record is thinner: the UK guideline on acute sore throat does not mention honey at all. This page walks through both records — the cough trials and their numbers, the throat data, what UK guidance recommends, and how the regulator judged the throat evidence.

Everything below is the published literature, cited as it comes.

Plate I · PaintingRecord

Does honey help a cough?

Probably — and the shape of the answer teaches as much as the answer itself. The anchor is a Cochrane systematic review (Oduwole and colleagues, CD007094, 2018 update): six randomised controlled trials, 899 children aged 12 months to 18 years, honey compared with dextromethorphan, diphenhydramine, salbutamol, bromelin, placebo and no treatment. Its headline finding, verbatim: “Honey probably relieves cough symptoms to a greater extent than no treatment, diphenhydramine, and placebo, but may make little or no difference compared to dextromethorphan.”

Behind that sentence are small differences on carer-scored symptom scales — parents rating how often and how badly their child coughed. Against no treatment, cough frequency differed by a mean of −1.05 (95% CI −1.48 to −0.62) and severity by −1.03 (−1.59 to −0.47), both from two studies in 154 children and both at moderate certainty — which in Cochrane’s grading means the reviewers think the true effect is probably close to this, while further research could still shift it. Against placebo, frequency differed by −1.62 (−3.02 to −0.22), from two studies in 402 children, and duration by −0.72 days (−1.31 to −0.13) — that last from a single trial of 102 children. The benefit is time-boxed in both directions: “Giving honey for up to three days is probably more effective in relieving cough symptoms compared with placebo or salbutamol”, while “Beyond three days honey probably had no advantage over salbutamol or placebo in reducing cough severity, bothersome cough, and impact of cough on sleep for parents and children (moderate-certainty evidence)”. On sleep quality salbutamol probably did better than honey.

Notice what was scored, because it defines what these trials can answer. The review’s outcomes were cough duration, symptomatic relief (frequency, severity and how bothersome the cough was), sleep for child and carer, quality of life, adverse effects, appetite and cost; phlegm, mucus clearance and expectoration are nowhere in that list, and sputum is named once, in the background, only to say that a cough may be “productive or unproductive of sputum”. The broader review below carries no phlegm outcome either — what it pools is combined symptom scores, cough frequency, cough severity, and recovery from throat pain and fever, and its one adult syrup trial recruited on non-productive cough — so the chesty version of the question has no trial answer here. What was measured is how often and how badly children coughed, and how everyone slept.

The limitations are the reviewers’ own, and they are specific. “Most of the children received treatment for one night, which is a limitation to the results of this review.” On harms the conclusion reads “There was no difference in occurrence of adverse events between the honey and control arms”, while the review’s own pooled figure for gastrointestinal symptoms against placebo is a risk ratio of 1.91 (95% CI 1.12 to 3.24) — roughly twice as likely on honey. Infants under 12 months were excluded throughout, honey’s use below that age being “restricted because of babies’ poor immunity against Clostridium botulinum”.

A broader systematic review and meta-analysis (Abuelgasim and colleagues, BMJ Evidence-Based Medicine, 2021) pooled 14 randomised trials, eight of them in children only, and it teaches a second lesson: the comparator decides how much a win means. Against “usual care” — a loose category the authors describe as showing “considerable diversity” — honey improved combined symptom score by a mean difference of −3.96 (95% CI −5.42 to −2.51), cough frequency by SMD −0.36 (−0.50 to −0.21) and cough severity by SMD −0.44 (−0.64 to −0.25), at moderate overall risk of bias, and the review’s stated conclusion is the affirmative one, on the record verbatim: “Honey was superior to usual care for the improvement of symptoms of upper respiratory tract infections”. The reviewers built that comparator on a stated premise — “There are no effective active treatments for URTIs, but many commonly used remedies. We therefore combined ineffective remedies as ‘usual care’” — and read the low heterogeneity of those comparisons as showing that “despite the variety of usual care treatments used by these studies, all were similarly ineffective”. Beating a basket of remedies the reviewers themselves class as ineffective is an easier test than beating a matched placebo.

The placebo comparison is where it gets genuinely interesting. Two trials, 372 patients, both at low risk of bias, pooled to SMD −0.63 with an interval of −1.44 to 0.18 — crossing zero — at 91% heterogeneity, because the trials disagreed. Which way they disagreed matters. The larger of the two, Cohen 2012 at 270 patients, estimated a clear benefit (SMD −1.03, −1.32 to −0.75); the smaller, Canciani, tested a honey-containing cough syrup rather than honey and found none (SMD −0.20, −0.59 to 0.19) — and in a sensitivity analysis that dropped the syrup, the one remaining study was significantly better than placebo. A third placebo trial, Waris 2014, could not be pooled at all, but reported honey reducing combined symptom score significantly more than placebo by the final day of the study (mean difference 3.99, p = 0.003). The authors’ own reading holds both halves at once: “Two of the three studies comparing honey to placebo indicated a beneficial effect of honey, but overall we do not have a strong evidence base from comparisons of honey against matched placebo” — and their Conclusions close that “further high quality, placebo controlled trials are needed”.

What does UK guidance recommend?

Honey is on the list — first, in fact. NICE’s guideline on acute cough (NG120) places it in a self-care list introduced like this: “Be aware that some people may wish to try the following self-care treatments, which have limited evidence of some benefit for the relief of cough symptoms:”. The first bullet is “honey (in people aged over 1 year)”, alongside pelargonium, guaifenesin and non-codeine cough suppressants for people aged 12 and over. The guideline’s evidence summary rests that bullet on an earlier version of the Cochrane review — three trials, 568 children and young people — and adds its own caution: “Honey should not be given to children until they are aged over 1 year because of concerns about infant botulism. It is also a sugar, and there are concerns about tooth decay.”

The NHS cough page takes the same view in plainer words. It lists “hot lemon and honey, which can have a similar effect to cough syrup (not suitable for babies under 1 year old)”, then qualifies it in the next line: “But there’s limited evidence to show that hot lemon and honey and herbal medicines help ease a cough.” Neither systematic review above separated the lemon, or the hot drink itself, from the honey; the same page adds that hot drinks should not be given to small children.

Fig. I · Two guidelines, facingDiagram

UK guidance on the two illnesses — the self-care lists, as this page quotes them

NG120 NICE guideline

Cough (acute): antimicrobial prescribing

“Be aware that some people may wish to try the following self-care treatments, which have limited evidence of some benefit for the relief of cough symptoms:”

The honey position · first bullet“honey (in people aged over 1 year)”

The rest of the list

• pelargonium
• guaifenesin
• non-codeine cough suppressants — people aged 12 and over

What the bullet rests onAn earlier version of the Cochrane review — 3 trials,568 children and young people.

“Honey should not be given to children until they are aged over 1 year because of concerns about infant botulism. It is also a sugar, and there are concerns about tooth decay.”

NG84 NICE guideline

Sore throat (acute): antimicrobial prescribing

No comparable invitation to try things is quoted here: what this page records of NG84 is the self-care advice its recommendations contain.

The honey position · nothing hereNo entry. NG84 does not mention honey anywhere in its recommendations.

The whole of the list

• paracetamol or ibuprofen
• adequate fluids
• medicated lozenges some adults may wish to try, which “may only help to reduce pain by a small amount”

The cough evidence does not carry acrossSore throat was not an outcome in the Cochrane review at all, which measured cough frequency, severity and duration, sleep quality and adverse effects.

The NHS self-care pages split the same way

CoughLists “hot lemon and honey, which can have a similar effect to cough syrup (not suitable for babies under 1 year old)”, then: “But there’s limited evidence to show that hot lemon and honey and herbal medicines help ease a cough.”

Sore throatDoes not mention honey either — it lists salt-water gargling for adults, fluids, cool or soft foods and rest.

That is public-health guidance about an ordinary self-limiting illness. It is not a health claim, and a food business may not repeat it as one.

Does honey help a sore throat?

Much less clearly — and the cough evidence does not carry across, because sore throat was not an outcome in the Cochrane review at all, which measured cough frequency, severity and duration, sleep quality and adverse effects.

The UK guideline on acute sore throat (NG84) does not mention honey anywhere in its recommendations: its self-care advice is paracetamol or ibuprofen, adequate fluids, and medicated lozenges some adults may wish to try, which “may only help to reduce pain by a small amount”. The NHS sore throat page does not mention honey either; it lists salt-water gargling for adults, fluids, cool or soft foods and rest.

What throat-specific data exist sit inside the Abuelgasim review’s adult analyses, and they found little. Four of the 14 trials had adult populations; the only pooled adult comparison available was honey versus usual care for cough frequency, SMD −0.19 (95% CI −0.47 to 0.09) — not significant. One adult trial (Nanda 2017) found no difference at five days in throat congestion (OR 0.73, 95% CI 0.42 to 1.27) or throat pain (OR 0.75, 0.43 to 1.32). Another (Gupta 2016) reported more patients reaching at least 75% improvement in throat irritation by day four, but tested a honey-based syrup containing herb extracts, not honey alone.

A separate literature covers the throat after surgery, and it leans the other way. A systematic review and meta-analysis (Hosseini and colleagues, 2025) pooled seven randomised trials, 710 children, of honey applied locally after tonsillectomy alongside standard treatment: pain intensity fell by a weighted mean difference of −0.90 (95% CI −1.32 to −0.48), at 92.5% heterogeneity and low certainty of evidence, the authors concluding that “cautious use is advised due to the limited quality of evidence”. Post-surgical pain in a hospital population is a different question from the sore throat that arrives with a cold.

Is manuka honey better for a cough or sore throat?

The one trial that tested it found no difference — and it was not testing a cold. None of the six trials in the Cochrane review used manuka: they used buckwheat, eucalyptus, Labiatae and citrus honeys, an unnamed honey from a village in Iran, and “the darkest locally available honey” in Kenya. The trial that did use it belongs to the surgical literature above. The BEE PAIN FREE trial — described by its authors as a multicentre, double-blind, randomised controlled trial, run across three centres in Western Australia — allocated 400 children after tonsillectomy to standard treatment alone, Marri honey, Manuka honey or placebo, and reported that “Treatment with honey at this frequency did not impact postoperative pain scores significantly, with all groups showing similar trajectories”; the authors record that children took fewer doses than the protocol asked and that 20% were lost to follow-up or withdrew. Beyond that trial, searches of the indexed literature return no randomised controlled trial of manuka honey for acute cough, or for the sore throat of an ordinary upper respiratory infection, in people. Manuka grading measures a chemical marker rather than any clinical outcome; the manuka evidence page reads that record on its own terms.

How did the regulator judge “soothing for the throat”?

Guidance to the public and authorised claims are different instruments answering to different tests, and honey is the cleanest example of the gap — which makes the regulator’s verdict worth reading closely.

Fig. II · The register rowDiagram

Great Britain nutrition and health claims register

The authoritative list for England, Scotland and Wales; only claims on it may be used in commercial communications about food.

FoodHoney, including the […] in honeyFood column abridged — full wording in the text below

Claim“Soothing for the throat”

StatusNon-authorised

Reason given

“Non-compliance with the Regulation because on the basis of the scientific evidence assessed, this food is not sufficiently characterised for a scientific assessment of this claimed effect and the claim could not therefore be substantiated.”

Underlying opinion, as the register cites itEFSA Journal 2010;8(2):1484

8 honey entries in the register — every one non-authorised. This is one of them.

Cough · the register lineNo entry. The register contains no honey entry for cough at all.

NICE may list honey among things people might try for a cough and the NHS may mention it, while a seller of honey has no authorised claim available.

The Great Britain nutrition and health claims register is the authoritative list for England, Scotland and Wales; only claims on it may be used in commercial communications about food. It holds eight honey entries, every one non-authorised, and one is the wording in question. The food reads “Honey, including the antioxidants in honey”, the claim “Soothing for the throat”, the status “Non-authorised”, and the reason “Non-compliance with the Regulation because on the basis of the scientific evidence assessed, this food is not sufficiently characterised for a scientific assessment of this claimed effect and the claim could not therefore be substantiated.” The register cites the underlying EFSA opinion as EFSA Journal 2010;8(2):1484.

Read that rejection carefully: it does not say honey does nothing. It says honey is too variable a food for a claimed effect to be pinned to it — one jar is not chemically the same as the next, so no single “honey” could be assessed. The register contains no honey entry for cough at all, which is why NICE can list honey among things people might try while the claims register, answering a stricter question with a stricter test, authorises nothing. The wider honey evidence page sets out the other seven rejected entries.

Who should not have honey, and when should a cough be looked at?

One rule is absolute: no honey for babies under 12 months. NHS guidance states: “Occasionally, honey contains bacteria that can produce toxins in a baby’s intestines, leading to infant botulism, which is a very serious illness. Do not give your child honey until they’re over 1 year old. Honey is a sugar, so avoiding it will also help prevent tooth decay.” That applies to every honey, however graded or processed.

Two further points deserve their own space. The first is allergy. A review of the allergic risks of bee products reports that “Allergies to honey are rare, and if they happen, they are often connected to sensitisation to pollen from the Asteraceae family or bee-derived proteins”, and that “Although rare, anaphylaxis can happen after honey consumption” — the allergens it names are the pollens carried in the honey and proteins of bee origin, and it gives no population prevalence figure. Anyone who has reacted to honey, to pollen or to other bee products has a question for a GP or pharmacist, not for a web page. The second is that honey is a sugar, which is why both NICE and the NHS attach a tooth-decay caution to it, and why an illness changes nothing about how it counts for anyone managing blood glucose — a matter for their GP or diabetes team.

The NHS puts the usual course of an acute cough at three to four weeks and most sore throats at about a week. A cough or sore throat that does not settle, or that comes with symptoms that worry you, is a matter for a GP or a pharmacist.

The short version

For acute cough in children over one, honey has real randomised evidence: a probable benefit over no treatment and placebo, bounded at roughly three days, at moderate certainty, and summarised by the reviewers themselves as no strong evidence either way. UK guidance lists it as something people may wish to try. For sore throat the record is thin and the guidelines are silent; the one manuka trial found no difference; and no honey before a first birthday, ever. That is the file — modest, specific, and worth knowing precisely.