Dossier
Margin
The short version · you have symptoms, you're at the chemist
Buy this. Skip this.
Buy
Treats the symptom you actually have.
Skip
The evidence says these do nothing for a cold.
Overpriced
Might do something. Not for that money.
If you only remember one thing: the effective decongestant is behind the counter and needs ID, and the one on the open shelf is the one that doesn't work. Prices are indicative Australian retail from July 2026 and will drift. None of this is medical advice, and pregnancy, blood pressure, glaucoma or heart conditions all change the answer, so ask the pharmacist who's already standing there.
Ranked by measured effect
Nearly all of this evidence is about colds, not flu. The trials below study the common cold and mixed upper respiratory infections, not laboratory-confirmed influenza. Suspected flu is a different clinical question, and a cold result should not be carried across to it.
Same product, different job. Vitamin C is close to useless at stopping you getting sick, and mildly useful once you already are. Those are two different questions and the packaging almost never separates them. Pick the job you're actually buying for.
Read the bars carefully
- Probiotics have the best prevention signal of anything in the aisle, which isn't saying much. Cochrane rates the evidence low to moderate quality with industry funding bias.
- The 50% Vitamin C result is real but it comes from soldiers and endurance athletes under extreme physical stress. It does not transfer to a normal winter.
Studied overseas, not on sale here
This is the result people quote when they say zinc shortens a cold. It is kept out of the ranking above because the product it describes is not something you can buy in an Australian pharmacy.
Why the zinc bar is not a recommendation
- You cannot buy the thing that was tested. The trials used zinc acetate lozenges. Australian pharmacy zinc is gluconate or a chelate, and usually a tablet or chewable rather than a lozenge, which matters because the proposed mechanism is zinc released slowly in the throat. The only zinc acetate lozenge we could find selling to Australia was out of stock, and its own label caps at 37.5mg a day, half the studied dose.
- The 2.37-day figure is a pooled estimate across 8 trials that mixed lozenges with intranasal zinc, at I²=97%. That is near-total disagreement between trials, and Cochrane rates the certainty low. Treat it as the best available guess, not a reliable number.
- The popular claim that only zinc acetate above 75mg a day works, and that citric acid kills the effect, comes from a separate re-analysis by Hemilä and Chalker and is disputed. No trial has directly tested whether formulation moderates efficacy.
- Harms are the better-evidenced half. Non-serious adverse events RR 1.34 (95% CI 1.15 to 1.55) at moderate certainty, against a low certainty benefit. Expect a metallic taste and possible nausea. This is about lozenges; intranasal zinc has been linked to lasting loss of smell and should be avoided.
- Elderberry, Pelargonium and olive leaf all claim duration benefits. Every one of those claims traces back to small, mostly manufacturer-funded trials that rigorous reviews rate very low quality, so they get no bar rather than a flattering one.
You think it works.
It doesn't.
Oral phenylephrine, the whole “PE” shelf
Around 38% of it survives your gut, and trials consistently find it no better than placebo for a blocked nose. The US FDA formally proposed removing it from approved decongestants in November 2024.
Symptom relief SourcedVitamin C, for not getting sick
29 trial comparisons, 11,306 participants, relative risk 0.97 (95% CI 0.94 to 1.00). That's a null result. It's the single most widely held belief in the category and it isn't true.
Prevention SourcedClaratyne for a cold
Second-generation antihistamines target allergic rhinitis. A cold isn't histamine driven, so there's nothing for them to do.
Symptom relief SourcedThe AUST L number meaning “approved”
It means the ingredients came off a permitted low-risk list and the sponsor must meet manufacturing and regulatory obligations. Nobody at the TGA checked whether the product works.
SourcedWouldn't think it works.
It does.
Honey, out of the supermarket
Cochrane found it statistically comparable to dextromethorphan for children's night cough, and better than diphenhydramine, at low certainty over short follow-up.
Symptom relief SourcedSalt water up your nose
Modest and low certainty. The measured change was −0.33 on a four-point scale, which the trial authors themselves call possibly minimal. Its real appeal is a very high safety margin.
Symptom relief SourcedProbiotics, of all things
A 24% relative reduction in people having at least one acute respiratory infection (RR 0.76), and 42% fewer antibiotic prescriptions. Low certainty, strain specific, and a 2026 analysis found no significant effect on incidence, so hold it loosely.
Prevention ContestedDoing almost nothing
The honest ceiling for this whole category is feeling less bad while a cold runs its course. Once you accept that, the cheap single ingredients look a lot better than the branded bundles.
OpinionThe worst deal in the aisle
We're more certain it causes side effects than that it helps
Codral, Lemsip, Demazin and every other all-in-one pack that puts a painkiller and a decongestant in the same tablet. Across 7 trials and 2,575 people, Cochrane measured how often they help and how often they cause a side effect. The harm is the better-evidenced of the two.
Out of 100 people who take it
21 helped
21 get a benefit they would not have had otherwise. Plenty of the other 79 still feel better, because colds improve on their own and placebo works, but not because of the tablet.
Moderate certaintyOut of 100 people who take it
6 harmed
6 get a side effect they would not have had otherwise. A smaller number than the benefit, but the better evidenced of the two, and that is the part worth sitting with.
High certaintyCochrane's own wording is that the effects on individual symptoms are “probably too small to be clinically relevant”, and that there's no evidence of effectiveness in young children at all.
The label that looks like an approval
AUST L is not AUST R
Armaforce, and most vitamins, herbals and immune-support formulas.
- Efficacy checked by the TGA
- No
- Who vouches for it
- The sponsor
- When it is checked
- Only if audited
The ingredients come off a permitted low-risk list. Nobody reads the evidence unless the product is later audited.
Paracetamol, ibuprofen, pseudoephedrine, benzydamine.
- Efficacy checked by the TGA
- Yes
- Who vouches for it
- The TGA
- When it is checked
- Before sale
Not a promise of a large effect. Older and generic registrations can rest on published literature or an established monograph.
Both numbers are printed in the same typeface, in the same corner of the box. Only one is a statement about whether the thing works.
What the TGA actually found on Armaforce’s main ingredient
- 1,368
- adverse event reports for Andrographis medicines to 31 Dec 2025. 859 name Armaforce.
- 287
- of those were anaphylaxis. 188 name Armaforce, and one person died in 2024.
- 1st
- the most-reported reaction is losing your sense of taste, ahead of anaphylaxis.
- 30min
- how quickly symptoms usually start. Most people affected had no history of allergy or asthma, and a viral infection may raise the risk, which is when you would be taking it.
- 0
- published figures for how many people take it. Without a denominator, none of the above converts into your personal odds.
- 2019
- warnings have been required since then and were strengthened in 2024. The TGA says they have not meaningfully reduced reports, and that stronger warnings, formulation limits and education were all assessed and none worked.
Consultation on removing the herb from low-risk listed medicines closed 8 May 2026 and was unresolved at the time of writing. The TGA has said cancellation and recall are possible “if warranted”. ARTG number verified against the current record; one research run reported 373050, which the record does not support.